SMOKING AND PERIODONTAL DISEASE D.F. Kinane* I.G. Chestnutt

Periodontology and Oral Immunology, University of Glasgow Dental Hospital and School, 378 Sauchiehall Street, Glasgow, G2 3JZ, Scotland, UK; *corresponding author ABSTRACT: Numerous investigations of the relationship between smoking and periodontal disease have been performed over the last 15 years, and there now exists a substantial body of literature upon which this current review is based. From both cross-sectional and longitudinal studies, there appears to be strong epidemiological evidence that smoking confers a considerably increased risk of periodontal disease. This evidence is further supported by the data emanating from patients who stop smoking. These patients have levels of risk similar to those of non-smokers. Numerous studies of the potential mechanisms whereby smoking tobacco may predispose to periodontal disease have been conducted, and it appears that smoking may affect the vasculature, the humoral immune system, and the cellular immune and inflammatory systems, and have effects throughout the cytokine and adhesion molecule network. The aim of this review is to consider the evidence for the association between smoking and periodontal diseases and to highlight the biological mechanisms whereby smoking may affect the periodontium. Key words. Smoking, cessation, nicotine, immunoglobulins, cytokines, periodontitis

(I) Introduction The concept that smoking tobacco may be prejudicial to periodontal health is not new, an association between acute necrotizing ulcerative gingivitis and smoking having been observed in the late 1940s (Pindborg, 1947). In the interim, numerous investigations of the relationship between smoking and the various other forms of periodontal disease have continued, and there now exists a substantial body of literature upon which this current review is based. The purpose of this review is: (1) to examine evidence for the association between smoking and periodontal disease; (2) to discuss possible biological mechanisms whereby smoking may adversely affect the periodontium; and (3) to consider the impact of smoking on periodontal treatment.

(11) Epidemiology (A) STRENGTH OF ASSOCIATION The stronger an association between a given factor and a disease, the more likely this factor will implicated as a risk factor. The strength of an association in both casecontrol and prospective studies can be measured by the relative risk, which is often expressed in terms of the odds ratio. Numerous cross-sectional studies on the effect of smoking on periodontal health have been reported, with odds ratios generally in the order of 2 to 6. One of the largest studies of risk factors for periodontal disease was that undertaken in Erie County, New York State. Involving 1361 subjects aged 25 to 74 years, this study showed that those who smoked were at greater risk for experiencing severe bone loss than those who did 356

not smoke, with odds ratios ranging from 3.25 to 7.28 for light and heavy smokers, respectively (Grossi et al., 1995). In a Swedish investigation of 155 patients with periodontal disease, a significantly higher percentage were found to be smokers than in the population at large, and the risk ratio was reported as 2.5 (Bergstrom, 1989). After controlling for confounding factors such as age, sex, plaque, and calculus, in a study of 615 American adults, the odds of having a mean probing depth of at least 3.5 mm in one randomly selected posterior sextant was reported as five times greater for smokers than for non-smokers (Stoltenberg et al., 1993). In 1991, a radiographic study of alveolar bone loss in Swedish dental hygienists demonstrated that the distance between the cemento-enamel junction and the interdental septum was significantly greater in smokers than in nonsmokers and increased with increasing smoking exposure. This study, in adults with good oral hygiene, suggested that smoking-related bone loss was not simply correlated with plaque levels (Bergstrom et al., 1991). A more recent study of 540 Swedish adults 20-70 years of age has revealed that the three variables smoking, greater age, and higher mean plaque levels were potential risk factors for severe periodontitis (Norderyd and Hugoson, 1998). A case-control study of the relationship between life-events and periodontitis has shown smoking to be statistically associated with periodontal disease, after controlling for oral health behavior and socio-demographic variables (Croucher et al., 1997). This work confirmed the earlier findings that periodontal disease experience is influenced by social and behavioral factors, and that smoking status was independent of other factors studied (Locker and Leake, 1993).

Oral Biol Med Crit Rev Crit Rev Oral Biol Med

(2000)

11(3):356-365 (2000)

by an array of studies with well-defined clinical criteria and more sophisticated data analysis (Mandel, 1994)1

In an investigation of behavioral and socio-demographic risk indicators of attachment loss in 873 subjects, aged at least 45 years, smoking was significantly associated with a higher probability of having one or more periodontal sites with severe attachment loss (Dolan et al. 1997). Smoking was found to be the strongest independent factor affecting periodontal status in 499 Finnish male industrial workers (Ahlberg et al., 1996). Although most studies report smoking to increase the risk of periodontal disease in the order of two- to sixfold, a study carried out in Northern Ireland suggested that the odds ratio for established periodontitis in a group of 82 young subjects (20 to 30 years of age) who regularly attended their dentist was as high as 14.1 (Linden and Mullally, 1994). Further evidence of the particularlv damaging effects of smoking on periodontal health in younger people was reported in a Swedish study. This longitudinal study investigated tooth loss in 273 individuals who were followed for a ten-year period. The risk attributable to smoking in younger males smoking more than 15 cigarettes per day was reported as 78% (Holm, 1 994 This is given further weight by the findings of Haber and colleagues, who, in determining the role of smokincg as a risk factor for periodcntitis in diabetic and non-diabetic study groups, suggested that in the nondiabetic group 51.% of the periodontitis in the 19-30year-olds and 32°o in those aged 31 to 40 years were associal-ed with smoking (Haber et c1l., 1993). Although the greatest number of investigations of the relationship between smoking and periodontal disease are c ross-sectional, a few longitudinal investigations have also been conducted. In a ten-year longitudinal radiographic study of alveolar bone loss which began in 1970, it was shown that in those subjects who had at least 2(0 teethi it the start of the study, smoking was a significant predictor of future bone loss (Bolin, 1986), and in a five-year study of attachment loss in 800 community-dwelling adults, smokers were found to be at an increased risk for attachment loss (Beck et al., 1997). A further longitudinal study, in which a wide range of clinical microbiological, and immunological indicators was correlated with disease progression, reported that over the one-year period of the investigation, smokers exhibited both greater attachment loss and bone loss when compared with their non-smoking counterparts. Smokers were shown to be at significantly greater risk for further attachment loss when compared with non-smokers, the odds ratio being quoted as 5.4 iMachtei et aii. 1997).

(C) DOSE-RESPONSE In determining attributable risk, it is of value to examine the relationship between the degree of exposure to an alleged risk factor and disease prevalence. The ability to demonstrate a dose-response strengthens the evidence of risk factor status. However, the absence of a doseresponse relationship does not necessarily rule out a causal relationship, since a threshold may exist above which disease develops. Grossi et al. (I1994) examined the relationship between smoking and attachment loss and demonstrated a dosedependent response, the odds for more severe attachment loss in smokers, compared with non-smokers, ranging from 2.05 for light smokers to 4.75 in heavy smokers. These findings support those of Alpagot and colleagues, who reported that probing depth was significantly correlated with 'packyears' (i.e., packs of cigarettes smoked per day multiplied by the number of years the subject has smoked) (Alpagot et cl., 1996). Furthermore years of exposure to tobacco products have been shown to be a statistically significant risk factor for periodontal disease in 1 156 community-dwelling New England elders, regardless of other social and behavioral factors ((ette et al. 19931. A Spanish survey involving 889 patients reported that gingival recession, pocket depth, and probing attachment level were significantly related to smoking status, and that attachment levels were proportionate to the quantity of cigarettes smoked. Smoking one cigarette per day, up to ten, and up to 20, increased probing attachment level by 0.5%, 500, and 10o, respectively. However, only in the latter group did loss of attachment differ significantly from that of non-smokers This led the authors to conclude that tobacco usage increases disease severity, and that this effect is clinically evident above a certain level of usage (Martinez-Canut et crl, 1 995). This suggestion of greater periodontal destruction above a certain level of smoking was suggested in a previous investigation of alveolar bone loss. Expressed as a percentage of tooth root length in 723 dentate adults, alveolar bone height was shown to be significantly lower in individuals smoking more than 5 g of tobacco per day compared with those smoking between I and 5 g of tobacco per day (Wouters et cl., 1993). Norderyd and Hugoson (1998) examined 547 Swedish adults and found that moderate to heavy smoking (greater thar or equal to 10 cigarettes per day) was associated with severe periodontitis, but that light smoking (fewer than 10 cigarettes per day) was not. The strong association found between smoking and advanced periodontitis is consistent with the hypothesis that smoking has cumulative detrimental effects on periodontal health (Horning et r1- 1992). Thus, there is good evidence that the more a patient smokes, the greater the degree of periodontal disease that will be experienced. Furthermore, there is a sug-

(B) CONSISTENCY OF OBSERVATIONS While many of the initial studies on the relationship

between smoking and periodontal disease reached conflicting and contradictory conclusions, in retrospect, such findings can be explained by the failure to account sufficiently for the multifactorial nature of periodontal disease. The ambiguity that clouded the early studies on the role of smoking ir periodontal disease has been resolved

5

0( ((33636

1(3).356-365 i20001

Gi

e

Oa

il

e

Crit Rev Oral Biot Med

357

gestion that this may worsen above a certain threshold. It is difficult to determine the strength of smoking as a risk factor, since a problem lies in accurate measurement of a subject's exposure to tobacco products over many years, and to date, most studies on the relationship between smoking and periodontal disease have determined smoking status by interview or questionnaire. However, it should be remembered that all retrospective studies are subject to recall bias, and while some studies have quantified lifetime exposure in packyears, current levels of smoking may not reflect past exposure. In one of the few studies which, to date, have measured cotinine, the severity of periodontal destruction, measured either as clinical attachment level or crestal bone height, was shown to be statistically positively correlated with serum cotinine levels (Gonzalez et al., 1996). Cotinine is the principle metabolite of nicotine and as such provides a valuable quantitative measure of smoking status. Patients' cotinine levels have recently been shown to correlate directly with outcomes of progressive periodontal breakdown (Machtei et al., 1997).

(D) SUBGINGIVAL

MICROBIAL FLORA

Analysis of the available data suggests that smokers may have a more 'pathogenic' microbial flora, in that Bacteroides forsythus was harbored subgingivally more in smokers than in non-smokers (Zambon et al., 1996). There was also a tendency for Porphyromonas gingivalis counts to be higher in smokers than in non-smokers, but this failed to reach significance. The data were adjusted for attachment loss and age, and thus the reported B. forsythus increase in smokers (former and current) is not simply because they have more severe periodontal disease. Similarly, current smokers have elevated Actinobacillus actinomycetemcomitans and B. forsythus counts, even after adjustments for periodontal disease and age. The finding of elevated levels of particular pathogens tends to vary across different studies. A more recent study using molecular analyses techniques for six putative periodontal pathogens found that current smokers displayed an increased risk for harboring Treponema denticola (OR = 4.61) (Umeda et al., 1998). Haffajee et al. (1997) have reported significant clinical improvements, following scaling and root planing (SRP), in subjects who had never smoked or who were past smokers, but not in current smokers. P. gingivalis, B. forsythus, and Treponema denticola were equally prevalent among current, past, and "never" smokers before therapy and decreased significantly post-SRP in all but the current smokers. Clinical improvement post-SRP in all patients was accompanied by a modest change in the subgingival microbiota, primarily reductions in P. gingivalis, B. forsythus, and T. denticola. Stoltenberg et al. (1993) studied periodontitis patients to determine if the prevalence of 5 bacteria commonly associated with periodontal disease differed between smokers and non-smokers. They studied P. gingivalis, A. actinomycetemcomitans, Prevotella intermedia, Eikenella

358

corrodens, and Fusobacterium nucleatum. No statistically significant difference in the prevalence of any of the bacteria was found between smokers and non-smokers. The investigators performed a logistic regression analysis on the data and found that smoking and mean probing depth > 3.5 mm were significantly associated with the presence of A. actinomycetemcomitans, P. intermedia, and E. corrodens (P < 0.05).

(E) CESSATION Further evidence of the role of smoking in periodontal disease comes from studies of patients who stop smoking. If smoking is associated with increased risk for periodontal disease, a reduction or elimination of tobacco use should reduce this risk and should be beneficial to the patient. There is good evidence that the prevalence of periodontal disease is less in former smokers than in those who continue to smoke. In a study comparing the prevalence of cigarette smoking among patients attending specialist periodontal and general dental practices, a dose-response was observed. After controlling for age and sex, the odds ratio for those who currently smoked compared with "never" smokers was 3.3, while former smokers compared with "never" smokers had a ratio of 2.1 with regard to the presence of moderate or advanced periodontitis (Haber and Kent, 1992). The prevalence and severity of periodontitis have been shown to be less in former smokers compared with current smokers, leading Haber to conclude that stopping smoking is beneficial (Haber, 1994). In a ten-year radiographic follow-up of alveolar bone loss, it was reported that progression of bone loss was significantly retarded in those who had given up smoking during the study, compared with those who continued to smoke (Bolin et al., 1993). Prospective observations of tooth loss in 248 women and 977 men, with a mean follow-up time of six years, indicated that individuals who continued to smoke cigarettes had in the order of 2.4- to 3.5-fold risk of tooth loss compared with non-smokers. The rates of tooth loss in men were significantly reduced after they quit smoking cigarettes, but remained higher than those in non-smokers. The authors concluded that stopping smoking significantly benefits an individual's likelihood of tooth retention, but it may take decades for the individual to return to the rate of tooth loss observed in non-smokers (Krall et al. 1997).

Summary Cross-sectional and longitudinal studies have indicated a strong relationship between smoking and increased risk for periodontal breakdown. Research performed in the last decade has proved beyond doubt that smokers have more periodontal problems than non-smokers. The evidence that smoking is a risk factor for periodontal disease is strengthened by the consistency of findings in different studies and in different populations, viz. North America, the Nordic countries, and the United Kingdom. There are

Med Oral Biol Rev Oral Biol Med Crit Rev

11(3):356

(2000)

11(3):356-365 (2000)

inconsistencies, however, in the studies reporting the most frequently detected pathogers in periodontal pockets of smokers. Pocket microflora differences between smokers and non-smokers appear to be largely studydependent, and little can be inferred from these studies on whether a specific microbial flora can increase the risk or be associated with periodontal disease in smokers. The strong epidemiological evidence that smoking confers an increased risk of periodontal disease is further supported bv evidence emanating from studies of patients who stopped smoking. The prevalence of periodontal disease is less in former smokers than in those who continue to smoke indicating a clear advantage to those stopping smoking in the benefit to periodontal health.

(111) Biological Effects of Smoking

in Periodontal Disease

Numerous studies of the potential mechanisms whereby smoking tobacco may predispose to periodontal disease have been conducted. A comprehensive review on the modifying effects of tobacco and smoking on the host immune and inflammatory response has been recently published (Barbour et al., 1997). This section will focus on the tissue effects of tobacco smoking peculiar to the periodontal tissues, and on the clinical conditions, and reviews the most periodontally relevant aspects of the host-microbe interactions influenced by smoking. The reader is referred to the in-depth host-response review by Barbour et'al. ( 1997) for the more specific and general systemic effects of smoking on the immune and inflammatorv systems.

(A) EFFECT OF NICOTINE ON THE PERIODONTAL TISSUES

Nicotine may cause a vasoconstriction in the peripheral blood vessels (Clarke et al., 1981) and thus may reduce the clinical signls of gingivitis. Evidence for this reduction in clinical disease expression comes from various sources, including Bergstrom (19901, who compared the compliance of smokers with that of non-smokers in an oral hygiene intervention program. The plaque index decreased in both groups, and, despite the similarity in plaque index, gingival bleeding was significantly lower in smokers than in non-smokers. These results suggest that, in smokers, the clinical expression of gingivitis (i.e., chronic inflammation) in response to plaque is suppressed. A stucdy with similar findings was conducted by Danielsen et al (1990), who set up an experimental gingivitis studv on smokers and non-smokers. Similar amounts of plaque accumulated in the two groups during the period of no oral hygiene, but clinically, smokers exhibited less gingival inflammation than non-smokers. A study by Holmes (1990) compared crevicular fluid flow in smokers and non-smokers with clinically healthy gingiva and the crevicular fluid flow of smokers in the areas physically exposed to smoke (maxillary lingual) and in areas not physically exposed to smoke (maxillary buccal). Smokers had significantly less crevicular fluid flow

than non-smokers. Interestingly, the exposed lingual areas of smokers showed no significant difference from the less exposed buccal areas, which suggests that the effect of nicotine may not be local or, if it is local, that it may be modified by the saliva and its effects dispersed. The authors suggest that the effect of tobacco smoke on clinically healthy gingiva may be through vasoconstriction rather than direct physical irritation. Kinane and Radvar 11997), investigating the responses of smokers and non-smokers, with and without subgingival antimicrobials, to instrumentation, also reported that gingival crevicular fluid (GCF) volumes were significantly lower among smokers than non-smokers. In this study, it was noted that, after therapy, the decrease in the GCF volume of smokers was less than that of non-smokers, regardless of treatment modality. However, the actual mean GCF volumes still remained lower in smokers than in non-smokers. These findings are consistent with a diminished peripheral blood flow leading to a diminished GCF flow.

(B) EFFECT OF NICOTINE ON THE IMMUNE AND INFLAMMATORY SYSTEMS Smoking has been shown to affect various aspects of the host immune response (AAP, 1996). Smoking may have an adverse effect on fibroblast function (Raulin et al., 1988), chemotaxis and phagocytosis by neutrophils (Kenney et al., 1977; Kraal et al., 1977), and immunoglobulin production (Holt, 1987; Johnson et al., 1990). To mount a successful response to the bacteria, immune cells must arrive at the inflammatory site in appropriate numbers. Nicotine increases intercellular adhesion molecule-I (ICAM-I) and endothelial leukocyte adhesion molecule-l (ELAM- ) on human umbilical vein cells (endothelial cells) and appears to increase soluble ICAM- I in the serum of smokers (Koundouros et al., 19961. These adhesion molecule changes may affect leukocyte binding to endothelial cells lining the capillaries and post-capillary venules and thus, may impede the recruitment of important host defense cells to the area of inflammation and microbial challenge. Alavi et al. (1995) compared elastase concentrations in the GCF from individual sites of smokers with those in the GCF of non-smokers. They found that smokers had lower elastase concentrations in GCF than did nonsmokers. The reasons for this are not immediately clear. Elastase is produced locally mainly by the polymorphonuclear leukocyte (PMN) cells and is not found in normal serum. Thus, vasoconstriction of blood vessels would not account for the decrease in GCF elastase concentration in non-smokers. It may be that PMNs are less functional (Wolff et al., 1994) or are present in reduced quantities in the gingival crevices of smokers due to the reduced vascularity of the region. (C) CYTOKINES AND SMOKING Recent reports suggest that host cytokine levels are influenced by smoking. Tappia et a1.) 11995) have shown that the

rtRe rlBo e 36-5(00 113

11(31-356-365 (2000)

Crit Rev Oral Biol Med

5

359

plasma responses of smokers following lipopolysaccharide stimulation differed from those of non-smokers, in that smokers had significantly more tumor necrosis factor alpha (TNFa-) and Interleukin-6 (IL-6) and also the acutephase protein ox2-macroglobulin. Bostrom et al. (1998b) have reported that smokers had significantly higher TNFx levels in their GCF than did non-smokers in untreated and treated periodontitis patients (Bostrom et al., 1998a,b). Kuschner et al. (1996) have reported a dose-dependent effect of smoking on IL- 1, IL-6, IL-8, and monocyte chemotactic protein (MCP)- I levels. In contrast to these studies, Madretsma et al. (I1996) have suggested that nicotine exerts a negative immunoregulatory effect through modulation of the cytokine production of mononuclear cells. The inhibitory effects of nicotine on IL-2 and TNFot production are similar to those seen for prednisolone (an established antiinflammatory steroid). In this study, the nicotine concentrations giving these anti-inflammatory effects fell within the normal plasma nicotine range of smokers. A further study by Bernzweig et al. (1998) found that levels of IL-113 from gingival mononuclear cells were decreased when these cells were exposed to nicotine. This last study, however, used unrealistically high levels of nicotine, approximately 75 times that of the mean saliva level of nicotine found in smokers. These in vitro studies on nicotine's effects on cell cytokines and the cytokine levels in patient smokers appear to be in conflict. These differences need to be investigated by careful clinical and laboratory studies which also consider that smoking's effects are more complex than just increasing the nicotine concentration, and that the concentrations of nicotine and the other chemicals and noxious stimuli related to smoking should be factored into the studies. These studies may have importance, given the weight currently being given to the theories that cytokine overproduction may be a detrimental host response which predisposes an individual to periodontitis (Kornman and di Giovine, 1998).

(D) SMOKING AND THE HUMORAL IMMUNE RESPONSE Several recent papers and reviews have suggested that existing risk and prevention models would be strengthened by the inclusion of data on host-defense mechanisms (Genco, 1992; Page, 1992; Haber, 1994; Quinn et al., 1996). Macrophages play important roles in both cellmediated and humoral immunity as antigen-presenting cells. However, antigens are presented in the context of class 11 major histocompatibility complex (MHCII) surface molecules. It has been shown that alveolar macrophages from smokers exhibit reduced expression of class II MHC (Pankow et al., 1991; Mancini et al., 1993). This may eventually lead to a reduction in the humoral immune response to invading organisms. Smoking has been shown to reduce the concentration of serum IgG generally (Ferson et al., 1979; Gulsvik and Fagerhol, 1979; Anderson et al., 1982; Hersey et al., 1983; Robertson et al., 1984; McSharry et al., 1985). Interestingly, seroconversion following hepatitis B vaccine occurs much more slowly in 360

smokers than in non-smokers, and the frequency of subjects undergoing an immune response is lower for smokers (Struve et al., 1992; Roome et al., 1993). More specifically, smokers have been shown to have reduced titers of serum IgG to P. intermedia and F. nucleatum (Haber, 1994). Quinn et al. (1996) have recently demonstrated that smoking tends to limit the production of IgG2 in generalized early-onset periodontitis (GEOP) patients. This is significant in that this isotype is associated with the humoral immune response against carbohydrate antigens commonly found on oral pathogens (Ling et al., 1993). Moreover, it has recently been shown that the level of IgG2 against A. actinomycetemcomitans is lower in smokers than in non-smokers among EOP patients (Tangada et al., 1997). Gunsolley et al. (1997) have shown that smoking modifies the concentrations of some lgG subclasses in specific racial and diagnostic groups. In Blacks who smoked, those with chronic adult periodontitis had lower IgG1, while those with generalized early-onset periodontitis had lower IgG2. In White subjects, complex relationships between smoking and allotypic markers were noted, but no influence of periodontal diagnosis was found. Thus, in addition to immunoglobulin allotype, smoking in Blacks appears to influence their IgG subclass concentrations. Thus, smoking may alter the type of humoral immune response seen.

Summary It is suggested by several studies that smokers may have more disease because their microbial flora may be more 'pathogenic'. However, host-related factors must influence this subgingival niche and the microflora therein. The host defense system will include the soluble and cellular components of the inflammatory and immune systems as well as the innate immunity afforded by such things as the epithelial barrier and fluid flow (GCF and saliva). The reduced GCF flow reported in smokers will mean that antibodies and other defense molecules derived from the serum will be reduced in quantity. A further consequence of reduced GCF flow would be fewer microbial nutrients and less flushing of the gingival crevice and removal of microbes and waste products as the GCF leaves the gingival crevice by the positive outward flow. These factors may all influence the flora at these sites. Smoking may affect the vasculature, the humoral immune system, and the cellular and soluble inflammatory system, and may have effects throughout the cytokine and adhesion molecule network. The relative importance of these smoking-related alterations and their precise mode of action in increasing the risk of periodontal disease remains to be elucidated.

(IV) Smoking and Periodontal Treatment (A) RESPONSE TO CONVENTIONAL TREATMENT The present literature suggests that the outcome in smokers is less favorable than in non-smokers. Ah et al.

Med Biol Med Oral Biol Rev Oral Grit Rev Crit

11(3)356-365

11(3):356-365 (2000)

(1994j have demonstrated a poorer response to periodontal treatment in smokers compared with non-smokers. Preber and Bergstrbm (I 990) found that, 12 months following surgery, smokers had a statistically significantly reduced probing depth reduction compared with nonsmokers, despite accounting for differences in levels of plaque accumulation. Preber et uil. ( 1995) studied the clinical and microbiological effects of non-surgical therapy and found that smokers had a less favorable outcome in terms of pocket depth reduction than did smokers. The study revealed no difference, however, between smokers and non-smokers in terms of the microbiological changes following therapy, i.e., the microflora was broadly similar in both categories of patients before and after treatment. Machtei et al. 1998) considered the changes in attachment level and alveolar bone levels approximately one year after the hygiene phase of therapy. Non-smokers had relatively stable bone height, whereas smokers exhibited an annualized rate of bone loss of 1. 17 mm. Furthermore, in a recent five-year study (Bostrom et al., 1998a), were found to exhibit less improvement compared with non-smokers in terms of bone height. Approximately 90% of patients who were categorized as having failed to respond to conventional therapy were smokers (MacFarlane et al., 1992; Wolff et al., 1994). Recent work by Colombo et l. (1998) has disagreed with this stated proportion, since these investigators found that only 25, of their patients were current smokers, but that 40%, were former smokers. Bostrom et al. (1 998a) suggested that former smokers often begin smoking again, and therefore one must interpret the status of the former smokers cautiously, since self-reporting of smoking status is not reliable (Gonzalez et al., 996). Although smokers will also benefit from treatment, albeit to a lesser degree, treatment failures tend to predominate among smokers. Kinane and Radvar (1997) found that the response to non-surgical mechanical therapy is particularly poor in deep pockets in smokers. Although the attachment gain was also greater among the non-smokers than the smokers, this was not significant. This indicates that, after treatment, a greater degree of recession occurred among the non-smokers compared with the smokers. In the description of the appearance of smokers' periodontal condition, and in studies looking cross-sectionally at smokers, a commonlv noted feature is the level of recession, which is often noted as worse in smokers than in non-smokers (Martinez-Canutt et a., 1995; Gunsolley et al., 1998).

11i3 56365200)

-20)0

Cit

sal inflammation mesial and distal to the implant. These observations were significantly higher in the maxilla of the smoking group than in the mandibles and in the maxilla of the non-smokers (p < 0.01). These findings suggest that implants placed in smokers have a greater risk of developing peri-implantitis than those in nonsmokers, particularly in the maxilla.

(C) EFFECTS OF STOPPING SMOKING ON PERIODONTAL TREATMENT The effect of stopping smoking on response to treatment is interesting. Kaldahl et al. (1996) have suggested that past smokers and non-smokers responded similarly to treatment. In addition, heavy smokers (over 20/day) had higher plaque levels during maintenance and the poorest response to treatment, but did not significantly differ from light smokers. A further study by Grossi et al. ( 1997) investigated the effect of cigarette smoking on patients' clinical and microbiological responses to mechanical therapy. Current smokers had less healing and reduction in subgingival B. forsythus and P. gingirzalis after treatment compared with former and non-smokers. These authors concluded that, since the healing and microbial responses of former smokers are comparable with those of nonsmokers, smoking cessation may restore the normal periodontal healing response. Clearly, stopping smoking is beneficial, but compliance with the smoking cessation regime is often dubious with these patients. Future studies should monitor the amount smoked by assaying serum cotinine, since patient compliance and selfreporting are unreliable (Gonzalez et al. 19961.

(D) HEALING IN SMOKERS Healing following conventional scaling and root planing is seen clinically as a reduction in pocket depth and is the result of a reduction in inflammation which causes tissue shrinkage or reduced inflammatory swelling and also an improved tissue tone. This improved tissue form is more resistant to pocket probing forces and is detected clinically as an increase in clinical attachment. The tissue shrinkage may result in recession which, together with the increased attachment, produces reduced probing depths. It has been hypothesized that, in smokers, much of the inflammatory tissue swelling before treat-

(B) RESPONSE TO MORE COMPLEX TREATMENT Tonetti et cr1L. 11995) performed a retrospective study that examined the effect of cigarette smoking on the healing responise following guided tissue regeneration (GTR) in deep .nfrabony defects. This study indicated that smoking was a significant factor in determining the clinical outcome. A risk-assessment analysis indicated that smokers had a -gnificantly greater likelihood than non-

I 3 )3556- 365

smokers of having a reduced probing attachment level gain following GTR. More recent studies have concurred with these findings (Cortellini et al., 1996. Trombelli and Scabia, 1997; Trombelli et al., 1997), and other investigators, when using regenerative procedures with allografts, have also found smoking to be detrimental to healing (Luepke et al., 1997; Rosen et al, 1996) Haas et al. (I 996) looked at the relationship between smoking and peri-implantitis. They found no significant difference between smokers' and non-smokers' mean maxillary and mandibular plaque indices. However, the smokers had higher bleeding indices and mean periimplant pocket depths, and greater peri-implant muco-

ev'Ora

Crit Rev Oral Biol Med

36

361

fibers as efficiently, and thus gingival tissue support and adaptation will be impaired or at least slowed, and poor tissue form will often result in greater microbial plaque retention around teeth. Another crucial cell of the dento-gingival barrier is the keratinocyte. Johnson et al. (1996) have shown that human gingival keratinocytes are induced to produce significantly increased amounts of IL-1 and prostaglandin E-2 (PGE2) by tobacco extracts. Furthermore, a constant feature of the junctional epithelium is the presence of PMN migrating through the epithelial layers and into the gingival crevice. The PMN is considered an important cell in the local host defense at this site. MacFarlane et al. (1992) reported that phagocytosis of polymorphonuclear leukocytes in refractory periodontitis patients was impaired. They found that 90% of these patients were smokers compared with 210% of the controls. No direct assessment of smoking or tobacco products on PMN function was assessed, however.

Summary In the treatment of smokers, the use of non-surgical therapy will generally be Figure. A diagram summarizing the interactions between smoking and other factors effective in terms of probing depth reducwhich could u[timately lead to periodontal destruction. tion and gingivitis, albeit to a lesser extent than with non-smokers. Similar ment may be absent, and thus this part of the post-therbenefits ma' y also pertain for surgical intervention strateapy tissue change may not contribute as much to the gies, but catution should be exercised in more advanced post-treatment pocket depth reduction in smokers comtypes of surf gery, e.g., regenerative surgery. Smoking influpared with non-smokers (Kinane and Radvar, 1997). All ences healirng, and therefore the capacity for regenerathings being equal, smokers could therefore have deeper tion, particuilarly bone regeneration, may be impeded. pockets after therapy than non-smokers, and these pockThe effects of smoking on the outcome of periodontal ets will continue to harbor quantitatively and qualitamae y be summarized as the following trends: therapy tively more pathogenic bacteria than shallower pockets. therm Coupled with this are the reduced fibroblast, PMN, and short-term eeffects in terms of less gingivitis resolution, less probing depth reduction, and less attachment gain epithelial cell function, reduced host defense response, in smokers. The longer-term effects are evidenced by the and reduced vascularity of the site. These tissue differences in smokers following the initial short-term healing fact that 80--90% of treatment failures occur in smokers after therapy (up to six weeks) may partly explain the difand 70-80% ' of implant failures occur in smokers. There is ferences in treatment response. now consideLrable evidence for the role of smoking in the Differences in healing in the medium to longer term etiology of p)eriodontal disease and its adverse influence (three months to one year) may be related to the cellular on the treat]ment of periodontitis (Fig.), such that advisand tissue differences in smokers and may account for ing patientss of the consequences of tobacco use is the refractory nature of the disease in many smokers and in the management of patients who smoke. essential the poorer longer-term healing response reported. Fibroblasts are an important cell in the periodontal healREFERENCE.S ing response, and Raulin et al. (1988) has reported an impaired function of fibroblasts in smokers. Fibroblasts Ah MK, Johrnson GK, Kaldahl WB, Patil KD, Kalkwarf KL have been shown to bind and internalize nicotine (Hanes (1994). Trhe effect of smoking on the response to et al., 1991), which may be detrimental to their function. periodonital therapy. J Clin Periodontol 21:91-97. Poorly functioning fibroblasts may not produce collagen Ahlberg J, Ttuominen R, Murtomaa H (1996). Periodontal 362

Crit Crit Rev Rev Oral Oral Biol Biol Med Med

l1(3):356-365

(2000)

11(3):356-365 (2000)

status among male industrial workers in southern Finland with or without access to subsidized dental care. Acta Odontol Scand 54: 166- 1 70. Alavi AL, Palmer RM, Odell EW, Coward PY, Wilson RF (1995). Elastase in gingival crevicular fluid from smokers and non-smokers with chronic inflammatory periodontal disease. Oral Dis 1: 10-1 14. Alpagot TI Wolff LF, Smith QT, Tran SD (1996). Risk indicators for periodontal disease in a racially diverse urban population. Clin Periodontol 23:982-988. American Academy of Periodontology (1996). Position paper. tobacco use and the periodontal patient. I Periodontol 07:5 1 -56. Anderson P. Pedersen OF, Bach B, Bonde G0 (1982). Serum antibodies and immunoglobulins in smokers and non-smokers. Clin Exp lniniunol 47:467-473. Barbour SE. Nakashima K, Zhang 1, Tangada S, Hahn CL. Schenkein HA, et al (1997). Tobacco and smoking: environmental factors that modify the host response (immune system) and have an impact on periodontal health, Crit Rev Oral Biol Med 8:4")7-460. Beck ID Cusmano L, Green-Helms W, Koch GG, Offenbacher S (1997). A 5-year study of attachment loss in community-dwelling older adults: incidence density. Periodont Res 32: 506- 5 1 5. Bergstrcim (1989). Cigarette smoking as risk factor in ,hroni cperiodontal disease. Comniunity Dent Orol Epidemiol 17 245-247. Bergstrom Ii 1990). Oral hygiene compliance and gingivitis expressi n in cigarette smokers. Scand I Dent Res 98:497-5Q3 Bergstr6m 1m Eliasson S, Preber H (1991). Cigarette smoand eriodontal bone loss. I Periodontol 62:242-246. king Bernzweig Ef Payne JB, Reinhardt RA, Dyer 1K, Patil KD 1998). Nicotine and smokeless tobacco effects on ging val an( peripheral blood -nononuclear cells. Clin Pertioloitol 25:246-252. Bolin A 1986. Proximal alveolar bone loss in a longitudiinal radiop_graphic investigation. Swed Dent 35(Supp l- 1-08. Bolin A Eklund G Frithiof L Lavstedt S (1993). The effect of changed smoking habits on marginal alveolar bone loss, A iongitudinal study. Swed Dent 1 17:211-216. Bostrom L. inder LE, Bergstrom I (1998a). Influence of smoking oni the outcome of periodontal surgery. A 5vear follow-up. Clin Periodontol 25:194-201. Linder LE, Bergstrorr (1998b). Clinical Bostrorm expressivr of TNF-alpha in smoking-associated periodontai disease. I Clin Periodontol 25:767-773. Clarke NG Shephard BC, Hirsch RS (1981). The effect of intra-arterial epinephrine and nicotine on gingival circulation. Oril Surq Oral Med Oral Pathol 52:577-582. Colombo AP Ettirmiadi C, Haffajee AD, Cugini MA, Socransky

6~20t 11( 36

II( 3 ):356-365 2000'.

SS (1998). Serum IgG2 level, Gm(23) allotype and FcgammaRlla and FcgammaRlllb receptors in refractory

periodontal disease. Clin Perodontol 25:465-474. Cortellini P, Paolo G, Prato P, Tonetti MS (1996). Longterm stability of clinical attachment following guided tissue regeneration and conventional therapy. Clin Periodontol 23:106-11 1. Croucher R, Marcenes WS, Torre MC, Hughes F, Sheiham A (1997). The relationship between life-events and periodontitis. A case-control study. Clin Peiodontol 24:39-43. Danielsen B, Manji F, Nagelkerke N, Fejerskov 0, Baelum V ( 1990). Effect of cigarette smoking on the transition dynamics in experimental gingivitis. Clin Periodontol 17:159-164. Dolan TA, Gilbert GH, Ringelberg ML, Legler DW, Antonson DE, Foerster U, et 1. (1997). Behavioural risk indicators of attachment loss in adult Floridians. J Clin Periodontol 24:223-232. Ferson MA, Edwards A, Lind GW, Milton GW, Hersey P (1979). Low natural killer-cell activity and immunoglobulin levels associated with smoking in human subjects. Int I Cancer 23:603-609. Genco RI (1992). Host responses in periodontal diseases: current concepts. I Periodontol 63:338-355. Gonzalez YM, De Nardin A, Grossi SG, Machtei EE, Genco RI, De Nardin E (1996). Serum cotinine levels, smoking, and periodontal attachment loss. J Dent Res 75:796-802. Grossi SG, Zambon II, Ho AW, Koch G, Dunford RG, Machtei EE, et al. (I1994). Assessment of risk for periodontal disease. 1. Risk indicators for attachment loss. J Periodontol 65:260-267. Grossi SG, Genco RI, Machtei EE, Ho AW, Koch G, Dunford R (1995). Assessment of risk for periodontal disease. 11. Risk indicators for alveolar bone loss. Periodontol 66:23-29. Grossi SG, Zambon I, Machtei EE, Schifferle R, Andreana S, Genco RJ et al. (1997). Effects of smoking and smoking cessation on healing after mechanical periodontal therapy. Am Dent Assoc 128 599-607. Gulsvik A, Fagerhol MK (1979) Smoking and immunoglobulin levels (abstract) Lancet 1 :449. Gunsolley 1C, Pandey JP, Ouinn SM, Tew Schenkein HA (1997). The effect of race, smoking and immunoglobulin allotypes on IgG subclass concentrations Periodont Res 32:381-387. Gunsolley IC, Quinn SM, Tew 1, Gooss CM, Brooks CN, Schenkein HA (1998). The effect of smoking on individuals with minimal periodontal destruction. Periodontol 69:165-170. Haas R, Haimbock W, Mailath G, Watzek G (1996). The relationship of smoking on pern-implant tissue a retrospective study. J Prosthet Dent 76 592-596. Haber J (1994). Cigarette smoking: a major risk factor for

ed36 ri Re OalBil 363 Crit Rev Oral Biol Med

periodontitis. Comnpend Contin Educ Dent 15:1002-1008. Haber 1, Kent R (1992). Cigarette smoking in a periodontal practice. I Periodontol 63:100-106. Haber J, Wattles J, Crowley M, Mandell R, loshipura K, Kent RL (1993). Evidence for cigarette smoking as a major risk factor for periodontitis. I Periodontol 64:16-23. Haffajee AD, Cugini MA, Dibart S, Smith C, Kent RL Jr, Socransky SS (1997). The effect of SRP on the clinical and microbiological parameters of periodontal diseases. J Clin Periodontol 24:324-334. Hanes PJ, Schuster GS, Lubas S (1991). Binding, uptake, and release of nicotine by human gingival fibroblasts. J Clin Periodontol 62:147-152. Hersey P, Prendergast D, Edwards A (1983). Effects of cigarette smoking in the immune system. Med I Austr 15:425-429. Holm G (1994). Smoking as an additional risk for tooth loss. I Periodontol 65:996-1001. Holmes LG (1990). Effects of smoking and/or vitamin C on crevicular fluid flow in clinically healthy gingiva. Quintessence Int 2 1:191-195. Holt PG (1987). lmmune and inflammatory function in cigarette smokers. Thorax 42:241-249. Horning GM, Hatch CL, Cohen ME (1992). Risk indicators for periodontitis in a military treatment population. J Periodontol 63:297-302. Jette AM, Feldman HA, Tennstedt SL (1993). Tobacco use: a modifiable risk factor for dental disease among the elderly. Am J Public Health 83:1271-1276. Johnson GK, Poore TK, Payne JB, Organ CC (1996). Effect of smokeless tobacco extract on human gingival keratinocyte levels of prostaglandin E2 and interleukin1. 1 Periodontol 67:116-124. Johnson JD, Houchens DP, Kluwe WM, Craig DK, Fisher GL (1990). Effects of mainstream and environmental tobacco smoke on the immune system in animals and humans. A review. Crit Rev Toxicol 20:369-395. Kaldahl WB, Johnson GK, Patil KD, Kalkwarf KL (1996). Levels of cigarette consumption and response to periodontal therapy. I Periodontol 67:675-681. Kenney EB, Kraal IH, Saxe SR, Jones 1 (1977). The effect of cigarette smoke on human oral polymorphonuclear leukocytes. J Periodont Res 12:227-234. Kinane DF, Radvar M (1997). The effect of smoking on mechanical and antimicrobial periodontal therapy. J Periodontol 68:467-472. Kornman KS, di Giovine FS (1998). Genetic variations in cytokine expression: a risk factor for severity of adult periodontitis. Ann Periodontol 3:327-338. Koundouros E, Odell E, Coward P, Wilson R, Palmer RM (1996). Soluble adhesion molecules in serum of smokers and non-smokers, with and without periodontitis. J Periodont Res 31:596-599. Kraal JH, Chancellor MB, Bridges RB, Bemis KG, Hawke JE 364

(1977). Variations in the gingival polymorphonuclear leukocyte migration rate in dogs induced by chemotactic autologous serum and migration inhibitor from tobacco smoke. I Periodont Res 12:242-249. Krall EA, Dawson-Hughes B, Garvey Al, Garcia RI (1997). Smoking, smoking cessation, and tooth loss. I Dent Res 76:1653-1659. Kuschner WG, D'Alessandro A, Wong H, Blanc PD (1996). Dose-dependent cigarette smoking-related inflammatory responses in healthy adults. Euro Resp J 9:1989-1994. Linden GI, Mullally BH (1994). Cigarette smoking and periodontal destruction in young adults. I Periodontol 65:718-723. Ling TY, Sims TI, Chen HA, Whitney CW, Moncla BI, Engel LD, et al. (1993). Titer and subclass distribution of serum IgG reactive with Actinobacillus actinomnycetemcomitans in localised juvenile periodontitis. Clin Imnmunol 13:101-112. Locker D, Leake IL (1993). Risk indicators and risk markers for periodontal disease experience in older adults living independently in Ontario, Canada. I Dent Res 72:9-17. Luepke PG, Mellonig JT, Brunsvold MA (1997). A clinical evaluation of a bioresorbable barrier with and without decalcified freeze-dried bone allograft in the treatment of molar furcations. J Clin Periodontol 24:440-446. MacFarlane GD, Herzberg MC, Wolff LF, Hardie NA (1992). Refractory periodontitis associated with abnormal polymorphonuclear leukocyte phagocytosis and cigarette smoking. I Periodontol 63:908-913. Machtei EE, Dunford R, Hausmann E, Grossi SG, Powell J, Cummins D, et al. (1997). Longitudinal study of prognostic factors in established periodontitis patients. Clin Periodontol 24:102-109. Machtei EE, Hausmann E, Schmidt M, Grossi SG, Dunford R, Schifferle R, et al. (1998). Radiographic and clinical responses to periodontal therapy. J Periodontol 69:590-595. Madretsma GS, Donze GI, van Dijk AP, Tak CJ, Wilson JH, Zijlstra FJ (1996). Nicotine inhibits the in vitro production of interleukin 2 and tumour necrosis factor-alpha by human mononuclear cells. Immunopharmacology 35:47-51. Mancini NM, Bene MC, Gerard H, Chabot F, Faure G, Polu IM, et al. (1993). Early effects of short-time cigarette smoking on the human lung: a study of bronchoalveolar fluids. Lung 171:277-291. Mandel 1 (1994). Smoke signals: an alert for oral disease. J Am Dent Assoc 125:872-878. Martinez-Canut P, Lorca A, Magan R (1995). Smoking and periodontal disease severity. I Clin Periodontol 22:743-749. McSharry C, Banham SW, Boyd G (1985). Effect of cigarette smoking on the antibody response to inhaled antigens and the prevalence of extrinsic allergic alveolitis among pigeon breeders. Clin Allergy 1 5:487-494.

Biol Med Crit Rev Crit Rev Oral Oral Biol Med

365

(2000)

11(3):356-365 (2000)

Norderyd 0 Hugoson A (1998). Risk of severe periodontal disease in a Swedish adult population. A cross-sectional study. Clin Periodontol 25:1022-1028. Page RC ( 1992) Host response tests for diagnosing periodontal diseases. Periodontol 63:356-366. Pankow W Neumann K, Ruscheff J, Schroder R, von Wichert P (1991). Reduction in HLA-DR density on alveolar macrophages of smokers. Lung 169:255-262. Pindborg II 1947). Tobacco and gingivitis. I. Statistical examination of the significance of tobacco in the development of ulceromembranous gingivitis and in the formation of calculus. Dent Res 26:261-264. Preber H Bergstrbm J (1990). Effect of cigarette smoking on periodontal healing following surgical therapy. J C/iin Periodontol 17:324-328. Preber H. Linder L, Bergstrbm 1 (1995). Periodontal healing and periopathogenic microflorai in smokers and nonsmokers. Clin Periodontol 22:946-952. Quinn SM. Zhang I-B, Gunsolley IC, Schenkein IG, Schenkeil-n HA, Tew JG (1996). Influence of smoking and race on immunoglobulin subclass concentratiOIcS in early-onset periodontitis patients. Infect Immnun 64 2500-2505. Raulin LA, McPherson JC, McQuade MJ, Hanson BS i1988). The effect of nicotine on the attachment of human fibroblasts to glass and human root surfaces in vitro. Periodontol 59:318-325. Robertson MD, Boyd JE, Collins HPR, Davis JMG (1984). Serum immunoglobulin levels and humoral immune competence in coalworkers. Arn Indust Med 6:387-393. Roome Al, WValsh SI, Cartter MIL, Hadler JL (1993). Hepatitis B vaccine responsiveness in Connecticut public safety personnel. I Am Med Assoc 270:2931 -2934. Rosen PS, Marks MH, Reynolds MA (1996). Influence of smoking on long-term clinical results of intrabony defects treated with regenerative therapy. J Periodontol 67 1159- 116W' Stoltenberg JL, Osborn IB, Pihlstrom BL, Herzberg MC, Aeppli DM, Wolff LF, et al. (1993). Association between -Jgarette smoking, ba terial pathogens, and

11(3) 356 365 12000) 1(13)-356-365 12000)

Rev Crit Crit Rev

periodontal status. J Periodontol 64:1 225- 1 230. Struve I, Aronsson B, Frenning B, Granath F, von Sydow M, Weiland 0 (1992). Intramuscular versus intradermal administration of a recombinant hepatitis B vaccine: a comparison of response rates and analysis of factors influencing the antibody response. Scand J Infect Dis 24:423-429. Tangada SD, Califano IV, Nakashima K, Quinn SM, Zhang J-B, Gunsolley IC, et al. (1997). The effect of smoking on serum IgG2 reactive with Actinobacillus actinomycetemcomitans in early-onset periodontitis patients. Periodontol 68:842-850. Tappia PS, Troughton KL, Langley-Evans SC, Grimble RF (1995). Cigarette smoking influences cytokine production and antioxidant defences. Clin Sci 88:485-489. Tonetti MS, Pini Prato G, Cortellini P (1995). Effect of cigarette smoking on periodontal healing following GTR in infrabony defects. A preliminary retrospective study. Clin Periodontol 22:229-234. Trombelli L, Scabbia A (1997). Healing response of gingival recession defects following guided tissue regeneration procedures in smokers and non-smokers. Clin Periodontol 24:529-533. Trombelli L, Kim CK, Zimmerman GJ, Wikesjb UM (1997). Retrospective analysis of factors related to clinical outcome of guided tissue regeneration procedures in intrabony defects. J Clin Periodontol 24: 366- 371. Umeda M, Chen C, Bakker 1, Contreras A, Morrison IL, Slots J (1998). Risk indicators for harboring periodontal pathogens. I Periodontol 69: 11 1 -1118. Wolff L, Dahlen G, Aeppli D (1994). Bacteria as risk markers for periodontitis. I Periodontol 65:498-5 1 0. Wouters FR, Salonen LW, Frithiof L, Hellden LB (1993). Significance of some variables on interproximal alveolar bone height based on cross-sectional epidemiologic data Clin Periodontol 20 199-206. Zambon IJ, Grossi SG, Machtei EE, Ho AW, Dunford R, Genco RJ (1996). Cigarette smoking increases the risk for subgingival infection with periodontal pathogens. I Periodontol 67:1050- 1054.

Oral Biol Med Med

Oral Biol

365 365