Risk Factors


The AAP Guidelines describe risk assessment as “…increasingly important in periodontal treatment planning and should be part of every comprehensive dental and periodontal evaluation.” Risk assessment goes beyond the identification of the existence of disease and its severity, and considers factors that may influence future progression of disease. It also helps predict a patient’s likelihood of developing the disease, and therefore improves clinical decision-making. The existing data, which is presented below, demonstrates a number of risk factors for progression of periodontal diseases. By using risk assessment for periodontitis, the clinician can focus on early identification and provide proactive, targeted treatment for patients who are at risk. It is hoped that the use of clinical risk assessment will become a routine part of a comprehensive examination for all periodontal/dental patients.

Nunn (2003) provides a good overview of some of the determinants associated with risk factors, as well as terminology to help with the understanding of this subject. Some of the risk factors outlined include: Age, Gender, Race, Smoking, Socio-Economic Status, Diet/Nutrition, Psychological Stress and Alcohol Consumption. Systemic factors include medication induced periodontal issues, obesity and diabetes. Additional factors include genetic factors, tooth-level factors and microbial factors. Nunn also reviews hierarchy of study design as a means of identifying the strength of evidence presented in association periodontal disease risk.

Heitz-Mayfield (2005) also provided a good review of individual predictive factors associated with a patient’s susceptibility to progression of periodontitis. The review reinforces the fact that cigarette smoking, high levels of pathogenic bacteria and poorly controlled diabetes present an individual with an increased risk of developing periodontitis.

In an attempt to provide clinicians with a simple tool to assist in determining the relative risk of an individual developing periodontal disease, Page et al (2002) developed the “Periodontal Risk Calculator.” This computer-based tool uses a mathematical algorithm to assign a relative weight to various risk factors based on information gathered from a routine periodontal examination. The scoring system was based on a 5-point scale and used such information as the patient’s age, smoking history, diabetic status, history of periodontal surgery, pocket depth, bleeding on probing and furcation involvement to name a few. The calculator was based on a retrospective analysis of 523 subjects with information gathered from a baseline examination. Subjects returned for reassessment in 3, 9 and 15 years for a reassessment and an update of the periodontal status. The results indicated a strong association between risk scores at baseline and the mean percentage of alveolar bone loss each year. There was also a strong correlation between baseline measurement and tooth loss. Those with a risk score of 5, for example, had a relative risk of 10.6 of losing a tooth over a 15-year period.

In a very large-scale epidemiologic study looking at risk factors for attachment loss in a United States population, Hyman et al (2003) did a cross-sectional study which included 12,325 adults, part of the NHANES III cohort of subjects. In the end, those with the greatest mean attachment loss were of male gender, generalized gingival bleeding, poor attendance for professional dental care and had a long history of cigarette smoking. In a smaller cohort study looking at 3 urban American minority populations, Craig and Haffajee et al (2001) recorded pocket depth, attachment level, gingival erythema, bleeding on probing, suppuration, presence of supragingival plaque, as well as smoking history in an Asian-American, African-American and Hispanic population. Although many confounders were noted, the study indicated that African-Americans had greater mean pocket depth, mean loss of attachment and a greater number of missing teeth relative to the two other ethnic groups.


Smoking has been a long-established risk factor associated with the development of periodontal disease. In 2004, Johnson and Hill aimed to review a number of aspects associated with the development and management of periodontal disease in a smoking population. In this article, the authors review the relevant history of research in this field. They review such issues as the strength of the association, consistency of findings, specificity of disease progression or cessation based on smoking status, as well as the biological gradient and plausibility, to name a few.

In 2000, Tomar and Asma reviewed the NHANES III data associated with smoking-attributable periodontitis. Although the limits of the NHANES III examination process have been documented elsewhere, the crude odds ratio between smoking and periodontitis was 3.58. Once adjusted for age, race, ethnicity, income level and education, the odds ratio rose to 3.97. There was also evidence of a dose-dependent relationship involving the number of cigarettes smoked, as well as the number of years smoking and the development of periodontal disease. The study suggested that 74.8% of periodontitis cases were attributable to smoking, with both current and former smokers contributing to this involvement.

In a study that reinforced the finding of Tomar and Asma, Susin et al (2004) attempted to estimate the severity of attachment loss attributable to cigarette smoking in a Brazilian population. 50% of the population studied had been exposed to cigarette smoking with a substantially higher attributable fraction in heavy smokers versus moderate smokers. In this study, the attributable fraction of disease in light smokers was similar to non-smokers. This was especially noteworthy in those 50 years of age and older. It was suggested that early smoking cessation programs might have helped to reduce periodontal disease activity.

The effects of smoking on periodontal therapy outcome have always been an important issue to discuss with patients. Heasman et al (2006) provided a review of the clinical evidence of treatment response in smokers versus non-smokers. Most of the studies reviewed were case-control, cross-sectional and parallel group studies of a short duration. The majority of trials reviewed showed significantly greater reduction in probing depths and bleeding on probing as well as a markedly greater gain of clinical attachment following both non-surgical and surgical treatments in non-smokers as compared with smokers. The differences are also seen in treatment of both Class I and Class II furcation defects. The review also suggested that there were benefits to quitting smoking in that, with limited data, the outcome of treatment in former smokers was superior to that of current smokers.

In a study focusing on healing following flap surgery, Scabbia et al., (2001) demonstrated that cigarette smoking negatively affects the healing response following periodontal flap debridement surgery. 57 patients were assessed and determined to have at least 3 teeth with similar periodontal involvement, necessitating flap debridement surgery. Thorough phase one treatment was completed prior to surgery. Measurements were made at baseline and 6 months post-surgery. Pre-surgical sites with > 4 mm PD were entered into the analysis. 28 smokers and 29 non-smokers were entered into the study. The average daily cigarette consumption was 21 + 9.8 cigarettes. Pre-surgical measurements were similar between the groups and improvements were seen in both smokers and non-smokers post-surgery. There was no statistically significant difference between the two groups. However, in deeper sites, non-smokers showed greater clinical attachment gain compared to smokers. 58% of sites in smokers had a decrease in probing depth to less than 3 mm, while 70% of non-smokers had a similar reduction in probing depth. 47% of sites with >7 mm probing depth pre-surgery reduced to less than 3 mm in a non-smoking population, while 16% of similar sites in a smoking population reduced to less than 3 mm. Therefore, it was concluded that smokers had a less favourable outcome following surgery compared to non-smokers, especially in deep pockets.

This evidence then suggests that there is definite benefit to smoking cessation plans in terms of improving the result of periodontal therapy. Krall et al. (2006) looked at the risk of tooth loss after cigarette smoking cessation. This prospective study of 789 men (part of the Veterans Administration Dental Longitudinal Study, 1968-2004) looked at tooth status and smoking status every 3 years for a maximum of 35 years. Three groups were assessed: 264 never smokers, 283 former smokers, and current smokers which were split in to “quitters” (129), those who quit smoking during follow-up and abstained from any type of tobacco product, and continuous smokers (113), those who continued to smoke cigarettes at each examination. The results indicated a 2.1 “hazard ratio” of tooth loss in current smokers, 2.0 in those with 1-year abstinence from smoking. A former smoker began to look like a non-smoker in terms of tooth loss after 15 years of abstinence. Therefore the results indicate that smoking cessation is beneficial for tooth retention, but long-term abstinence is required to reduce the risk to the level of people who have never smoked.

Palmer et al, in a 2005 Journal of Clinical Periodontology supplement, review the potential biological mechanisms underlying the effects of tobacco smoking on the development of periodontitis. The primary mechanisms suggest that differences in bacteria, blood flow within the gingiva, changes in neutrophil function, as well as cytokine activity changes and the effects of nicotine, provide the most relevant reasons for the differences observed. The review of studies suggested that smokers have a higher prevalence of Aa, Tf and Pg species of bacteria than non-smokers. T.denticola is also elevated in smokers. Current smokers tend to have a higher prevalence of periopathogens at shallow sites relative to non-smokers. Histologically, gingival blood flow significantly increased 3 days following smoking cessation and further increases occurred up to 4-8 weeks later. Smoking can also suppress neutrophil chemotaxis and increase the release of proteolytic enzymes that leads to cell death. Finally, studies show the detrimental effects of nicotine itself on fibroblast function, including fibroblast proliferation, PDL fibroblast adhesion to root surfaces and cytotoxicity.


See Oral-Systemic


Kornman et al (1997) discussed the concept of interleukin-1 genotype as a severity factor in adult periodontal disease. With the identification of a polymorphism evidenced by a 4-fold increase IL-1B production in those with a composite genotype IL1-A allele 2 -889 plus IL-1B allele 2 +3953, there was a 6.8 odds ratio of severe versus mild periodontal disease development. It was believed from this study that as association with increased IL-1 production might be a strong indicator of susceptibility to severe periodontitis in an adult population.

In 1999, McGuire et al as part of their series of articles looking at predictors of long-term outcome, looked at Il-1 genotype assessment as a means of assigning prognosis and predicting tooth loss. This study assessed 42 patients in maintenance care for 14 years. 16 of these patients tested positive for the gene associated with elevated IL-1 production. Within this population, nine subjects were current smokers and 30 had a history of smoking. Both a positive genotype and heavy smoking were significantly related to tooth loss. Positive IL-1 genotype increased the risk of tooth loss by 2.7 times and heavy smoking by 2.9 times. The combined effect of IL-1 genotype positive and heavy smoking increased the risk of tooth loss by 7.7 times. It was also determined that other clinical parameters such as tooth loss, mobility and probing depth did not add to the prognostic model established by the smoking/genetic component.

Michalowicz et al (2000) added the genetic evidence in a twin study that assessed probing depth, attachment level, plaque and gingival index on all teeth. It was determined that monozygotic twins were more similar than dizygotic twins in terms of clinical parameters and severity of disease. It was also determined that approximately half of the variance in disease in this population is attributed to genetic variance.

In a 2012, Karimbux et al provided a review of 27 studies that concluded there were significant clinical effects based on the variations seen in IL-1A and IL-1B production. What was also suggested was that these results were applicable to a white population of Northern European descent. The interleukin variations are not necessarily extrapolated to other populations (i.e., Asian population).

Finally, Socransky et al (2000) assessed the microbiological parameters associated with IL-1 gene polymorphism in periodontitis patients. In this study of 108 subjects, all subjects underwent a fingerstick blood sampling to assess for Il-1A (+4845) and IL-1B (+3954) genotyping using a PCR-based method. All subjects had up to 28 teeth sampled for bacterial deposits, which were assessed using the checkerboard DNA-DNA hybridization technique. Within the test population, 28 were considered genotype positive. Within these positive subjects, red and orange complex species, as well as S. intermedius and S. gordonii, were significantly higher than genotype negative subjects. It was surmised that the increased cytokine levels of the genotype positive subjects might provide a preferred environment for bacterial growth.


Madianos et al (2013) provided an overview of the physiological mechanisms associated with a normal pregnancy.  After conception, the placenta invades and grows as a result of support derived from maternal uterine tissue.  Through the vessel-rich placenta, there is exchange of nutrients and waste between the mother and fetus.  Having the necessary resources, the fetus grows in the amniotic fluid, which is contained by the amniotic sac.  As pregnancy progresses, amniotic fluid levels of prostaglandin E2 (PGE2) and inflammatory cytokines (TNF-a and IL-1b) rise steadily until a critical threshold level is reached to induce rupture of the amniotic sac membranes, uterine contraction, cervical dilation and delivery.  Thus, the inflammatory signalling controls normal parturition.  This process represents a triggering mechanism that can be modified by external stimuli infection and inflammatory stressors.

Adverse pregnancy outcomes can include any or all of the following:

Preterm Birth (PTB) - <37 weeks of gestation

Very Preterm Birth (VPTB) – 32 gestational weeks

Low Birth Weight (LBW) - < 2500 g at birth

Pre-eclampsia – Defined as persisting elevated diastolic blood pressure  (> 90 mg Hg), proteinuria (>300 mg in a 24-hour sample) and presence of edema.  Goldenberg et al (2000) reviewed a number of risk factors for preterm birth, which include:

Young maternal age

Multiple gestation

Small weight gain during pregnancy

Cervical incompetence

Smoking, alcohol, drug abuse

Black race

Maternal infection.

Yet despite all of these known risk factors, as much as 50% of the incidence of preterm birth remains unexplained by these factors alone.

One of the first published studies on this subject was that of Offenbacher et al (1996) in a case-control study examining 124 mothers.  93 “cases” gave birth to children with a birth weight of <2500 g and < 37 weeks gestation.  46 “controls” were mothers who delivered at term infants of normal birth weight.  The assessment looked at known obstetric risks including tobacco and drug use, alcohol consumption, level of prenatal care, parity, genitourinary infections, nutrition and finally periodontal attachment loss.  There was a small but statistically significant difference in attachment loss between cases and controls (3.1 mm to 2.8 mm), however when controlling for other risk factors it was determined that periodontal disease conferred an adjusted OR of 7.9 for preterm low birth weight babies.  Although this result was never replicated, other studies have shown a positive association between periodontitis and adverse pregnancy outcomes (Dasanayake et al 2003, Canakci et al 2004, Goepfert et al 2004, Mokeem et al 2004, Radnai et al 2004).  However, there were also a number of studies that failed to document this association (Davenport et al 2002, Buduneli et al 2005, Moore et al 2005).

A prospective study by Jeffcoat et al (2001) published in the Journal of the American Dental Association looked at 1313 women at 21 to 24 weeks of gestation.  In this study, more than 90% of all sites presented with attachment loss of 3 mm or more.  The Odds Ratio for early delivery of pregnant mothers with generalized periodontitis was as much as 7.07 for delivery before 32 weeks. 

Boggess et al (2006) looked at the possibility that periodontal disease may be associated with infants born that are small for gestational age.  This prospective study measured periodontal parameters at pre-natal visits, as well as C-reactive protein levels from maternal serum.  6.6% of subjects were delivered as small for gestational age.  14.3% of subjects had moderate to severe periodontal disease.  When adjusted for age, smoking, drugs marital and insurance status and pre-eclampsia there was a relative risk of 2.3 (1.1-4.7) of having a small for gestational age infant in association with the presence of periodontal disease.  Additionally, serum CRP levels were higher in women who had a small for gestational age infant.  It was felt that the risk may be mediated by an increase in the maternal systemic inflammatory load, which was in response to the challenge posed by the oral pathogens.

One of the early intervention studies examining the effects of periodontal therapy on adverse pregnancy outcome was done by Mitchell-Lewis et al (2001).  This case-control study presented a cohort with a rather high incidence of preterm low birth weight babies (16.5%).  Within this population of 164 women, 74 received oral prophylaxis during pregnancy, and 90 women received no pre-natal care.  Those having PTLB babies harboured significantly higher levels of T. forsythia and C. rectus with consistently elevated counts for a number of species examined.  In the end, the study found that 18.9% of the women who did not receive periodontal intervention had a PTLB baby, while 13.5% of cases occurred in those who did receive such therapy. 

Lopez et al (2002) provided evidence of improved pregnancy outcome in a high risk population following periodontal therapy.  400 pregnant women with periodontal disease in Chile were randomly assigned to a treatment group who received periodontal therapy before 28 weeks gestation.  The control group received treatment after delivery.  The incidence of an adverse pregnancy outcome was 1.8% in the treatment group and 10.1% in the control group, leading to an odds ratio of 5.5.  In a similar study performed on patients with pregnancy-associated gingivitis, Lopez et al (2005) looked at 870 pregnant Chilean women who were also randomly assigned to a treatment group prior to 28 weeks gestation.  The incidence of adverse pregnancy outcome in the treatment group was 2.1% and the control group (who also received treatment after their pregnancy) was 6.7%.  The odds ratio of reduced adverse pregnancy outcome was 3.3, which also demonstrated a positive effect of periodontal therapy on treatment outcome.

Finally, Michalowicz et al (2006) made an important contribution to this question with a multi-centered, blinded, control trial of non-surgical periodontal therapy on pregnancy outcomes.  823 women were randomly assigned to a treatment and control group.  Treated patients underwent scaling and root planning before 21 weeks gestation, plus additional recall visits which included monthly polishing and reinforcement of oral hygiene.  Treatment was provided on an as-needed basis until delivery.  The control group also received oral hygiene instruction during their pregnancy but did not receive scaling and root planning until after delivery.  The incidence of pre-term births (<37 weeks gestation) was 12.0% in the treatment group and 12.8% in the control group.  The study concluded that periodontal treatment during pregnancy was safe but did not significantly alter birth outcomes.  A meta analysis by Michalowicz et al in 2013 of the studies published up to 2012 reinforced the outcomes stated in the 2006 paper.


In 2012, the American Heart Association provided a scientific statement in the journal Circulation (Lockhart et al, 2012), which presented a review of the literature to determine if there was data supporting the notion of an independent association between atherosclerotic vascular disease (ASVD) and periodontal disease.  There are three theories about the possible link between cardiovascular disease (CVD) and periodontal disease: i) Indirect mechanism (systemic inflammation) ii) Indirect mechanism (mimicry) iii) Direct mechanism (vascular infections by periodontal pathogens). In terms of assessing systemic inflammation, it was noted that inflammatory markers are elevated in both PD and CVD (ie:  C-reactive protein, TNF-a, MMP’s, IL-6).  The overall inflammatory reaction leading to vascular tissue damage through B-cell and T-cell activity accounts for an indirect mechanism of “mimicry”.  Finally the evidence of periodontal pathogens adhering to vascular walls and being found in arterial plaques suggests a possible direct mechanism of association between CVD and PD.  However, from this assessment, a direct causal relationship has yet to be confirmed. 

Schenkein et al (2013) expands upon the relationship of the inflammatory mechanisms linking periodontal disease to cardiovascular disease.  This article tends to reinforce the review by Lockhart et al and suggests that although there are many common factors noted in patients with PD and CVD, including genetics, diet, and lifestyle, it remains difficult to draw a causal relationship between CVD and PD. 

Dietrich et al (2013) also provided a systematic review of the evidence for an association between CVD and PD.  12 studies were looked at in the review, all of which reported some positive association between periodontal disease and CVD.  The review suggested that there was a stronger association in younger adults relative to a population older than 65 years.  However, there was insufficient evidence for an association between periodontal disease and secondary cardiovascular events (patients with established CVD). 

Three major interventional trials regarding the treatment of periodontal disease and its effects on CVD outcomes (D’Aiuto 2004, Tonetti 2007, Offenbacher 2009) were summarized in the paper by D’Aiuto (2013).  This systematic review looked at non-surgical therapy effects on endothelial function and the level of inflammatory markers.  One consistent finding was the fact that a consistent reduction in C-reactive protein was noted following periodontal therapy.  Improved endothelial function was also noted following therapy.  Endothelial dysfunction is predictive of future CVD risk.  Although there was an improvement in these surrogate markers, overall no direct effects of periodontal therapy on reducing cardiovascular disease outcomes has been noted.   A “cause-effect” relationship has yet to be established.