Third Molars and Periodontal Pathology in American Adolescents and Young Adults: A Prevalence Study

J Oral Maxillofac Surg 68:325-329, 2010

Third Molars and Periodontal Pathology in American Adolescents and Young Adults: A Prevalence Study George H. Blakey, DDS,* Savannah Gelesko, BSPH,† Robert D. Marciani, DMD,‡ Richard H. Haug, DDS,§ Steven Offenbacher, DDS, PhD,储 Ceib Phillips, PhD, MPH,¶ and Raymond P. White, Jr, DDS, PhD# Purpose: To assess the association between visible third molars and the prevalence of periodontal

inflammatory disease of non–third molars. Patients and Methods: Subjects aged 14 to 45 years with 4 asymptomatic third molars were enrolled in an institutional review board–approved study. Subjects were classified based on whether at least 1 third molar was visible or all third molars were not visible. Full-mouth periodontal probing depth (PD) data, with 6 sites per tooth, were obtained as a measure of a subject’s periodontal status. At least 1 non–third molar PD of 4 mm or greater was indicative of periodontal inflammatory disease. Outcomes for the respective groups were compared by use of Cochran-Mantel-Haenszel row mean score statistics. The level of significance for differences was set at .05. Results: The 342 subjects in the visible group were significantly older, with a median age of 26 years (interquartile range, 22.4-32.2 years), as compared with the 69 subjects in the not visible group, with a median age of 21 years (interquartile range, 18.8-24.9 years) (P ⬍ .01). The proportion of males and females was not statistically different between groups (P ⬎ .05). Most subjects were white. Significantly more subjects with at least a college education were in the visible group than in the not visible group (P ⬍ .01). The rate of tobacco use was low and did not differ between groups. Subjects in the visible group were significantly more likely to have at least 1 PD of 4 mm or greater on non–third molars than those in the not visible group: 59% versus 35%. In both groups, first/second molars were more affected than nonmolars when we controlled for differences in age between groups. Conclusions: The visible presence of third molars in adolescents and young adults was significantly associated with periodontal inflammatory disease of non–third molars. © 2010 American Association of Oral and Maxillofacial Surgeons J Oral Maxillofac Surg 68:325-329, 2010 Almost 50 years ago, Ash et al1,2 suggested an association between the presence of third molars and periodontal pathology affecting adjacent molars. However, limited population data on the association

between the presence of third molars and periodontal inflammatory disease of non–third molars exist because third molar data are often not collected from subjects in clinical or population studies. To compli-

*Clinical Associate Professor, Department of Oral and Maxillofacial Surgery, School of Dentistry, University of North Carolina, Chapel Hill, NC. †Dental Student, School of Dentistry, University of North Carolina, Chapel Hill, NC. ‡Chief, Division of Oral and Maxillofacial Surgery, College of Medicine, University of Cincinnati, Cincinnati, OH. §Section Chief, Oral and Maxillofacial Surgery, Carolinas Medical Center, Charlotte, NC. 储Ora Pharma Distinguished Professor, Department of Periodontology, School of Dentistry, University of North Carolina, Chapel Hill, NC. ¶Professor, Department of Orthodontics, School of Dentistry, University of North Carolina, Chapel Hill, NC.

#Dalton L. McMichael Distinguished Professor, Department of Oral and Maxillofacial Surgery, School of Dentistry, University of North Carolina, Chapel Hill, NC. Address correspondence and reprint requests to Dr White: Department of Oral and Maxillofacial Surgery, School of Dentistry, University of North Carolina, CB 7450, Chapel Hill, NC 27599-7450; e-mail: [email protected]. © 2010 American Association of Oral and Maxillofacial Surgeons

0278-2391/10/6802-0015$36.00/0 doi:10.1016/j.joms.2009.04.123

325

326 cate the situation further, many third molars are removed in adolescents and young adults without documentation of the reason for removal.3 Elter et al4 have analyzed the association between the presence of at least 1 visible third molar, or lack thereof, and the periodontal health of an adjacent second molar in subjects aged 18 to 34 years enrolled in the Third National Health and Nutrition Examination Survey (NHANES III) of the US population conducted by the National Center for Health Statistics of the Centers for Disease Control and Prevention from 1988 to 1994. A visible third molar detected in a subject was significantly associated with finding at least 1 periodontal probing depth (PD) of 5 mm or greater on an adjacent second molar. In a multivariable model including covariates clinicians usually associate with periodontal disease, these odds were comparable to the odds of an association between current or former smoking and at least 1 periodontal PD of 5 mm or greater on an adjacent second molar. No other similar population data have been reported from adolescents and young adults. Moss et al5 studied a population of 7,000 older subjects with a mean age of 62 years. If a third molar was visible, a significantly greater number of periodontal probing sites with a clinical attachment level of at least 3 mm were detected on first/second molars, as compared with first/second molars in subjects with no visible third molars, after adjustment for ethnicity, gender, age, smoking, education, and income. Similarly, significantly more probing sites of at least 4 mm were detected on first/second molars in subjects with at least 1 visible third molar than in subjects with no visible third molars, after adjustment for covariates. The 2 population studies cited previously have suggested that the visible presence of a third molar increased the risk of periodontal inflammatory disease being detected on non–third molar teeth. However, the findings from these studies represent secondary analyses of data collected for other purposes, and neither study had access to radiographs to confirm the presence of third molars. The objective of this analysis was to determine whether the purported relationship could be confirmed in a sample of subjects enrolled in a longitudinal cohort study with documented data on the presence of 4 asymptomatic third molars.

Patients and Methods Healthy adolescent and young adult subjects with 4 asymptomatic third molars were recruited and voluntarily enrolled in an institutional review board– approved cohort study at 2 academic clinical centers, the University of Kentucky (Lexington, KY) and the University of North Carolina (Chapel Hill, NC), over a 4-year period ending in 2002. Inclusion criteria for the

THIRD MOLARS AND PERIODONTAL DISEASE

study dictated that subjects be healthy (American Society of Anesthesiologists I or II), be aged between 14 and 45 years, and have 4 third molars with adjacent second molars verified by clinical examination and panoramic radiographs. Subjects with the most severe form of periodontal disease (American Academy of Periodontology IV) or who had taken antibiotics within 3 months before enrollment were excluded from participation. Baseline data taken at enrollment from all subjects were used for this cross-sectional analysis. Demographic data and data assessing oral health were collected from each subject. Full-mouth periodontal PD data, with 6 sites per tooth, were obtained clinically at enrollment as a measure of a subject’s periodontal status. PDs were rounded to the lowest whole number (eg, a PD of 3.8 was rounded to 3.0). For these analyses, each subject was classified based on whether at least 1 third molar was visible (visible group) or all third molars were not visible (not visible group). Non–third molars in both groups were divided into segments: first/second molars (48 probing sites) and nonmolars (120 probing sites). At least 1 PD of 4 mm or greater was considered an indicator for the presence of periodontal inflammatory disease.6,7 The periodontal status of the 2 groups— defined by the presence of at least 1 PD of 4 mm or greater, or lack thereof, detected on at least 1 first/second molar or on nonmolars—was compared by use of the Cochran-Mantel-Haenszel ␹2 test after adjustment for age by stratification (⬍25 years or ⱖ25 years). The demographic differences between the 2 groups were compared to assess whether factors other than visible third molars identifiable at enrollment were related to the detection of non–third molar periodontal pathology. Cochran-Mantel-Haenszel row mean score statistics were used to compare age, and the Fisher exact test was used to compare the ethnicity, gender, and education level of the 2 groups; significance was set at .05.

Results Data obtained at enrollment from 411 subjects were available for our cross-sectional analyses (Table 1). The visible group consisted of 342 subjects (83%) who had at least 1 third molar exposed; the not visible group consisted of 69 subjects (17%) who had all 4 third molars not exposed. The proportion of males and females was not statistically different between groups (P ⬎ .05). Subjects in the visible group were almost evenly divided between females (50.7%) and males (49.3%). More females (62.3%) were represented in the not visible group than males (37.7%). Overall, the ethnic distribution of the visible and not

327

BLAKEY ET AL

Table 1. COMPARISON OF DEMOGRAPHIC CHARACTERISTICS OF SUBJECTS (N ⴝ 411): THOSE WITH AT LEAST 1 THIRD MOLAR VISIBLE (n ⴝ 342) (VISIBLE GROUP) AND THOSE WITH NO THIRD MOLARS VISIBLE (n ⴝ 69) (NOT VISIBLE GROUP)

Subjects With ⱖ1 Third Molar Visible (Visible Group) Gender Female Male Ethnicity* White African American Asian Hispanic Other Education High school graduate or less Some college College graduate or higher Use of tobacco products† Yes No Median age at enrollment (IQR) (y)

Subjects With No Third Molars Visible (Not Visible Group)

P Value .08

173 (50.7) 168 (49.3)

43 (62.3) 26 (37.7)

257 (75.3) 47 (13.8) 26 (7.6) 5 (1.5) 6 (1.8)

56 (81.1) 11 (15.9) 1 (1.5) 0 (0) 1 (1.5)

28 (8.2) 107 (31.4) 206 (55.1)

17 (24.6) 28 (40.6) 24 (34.8)

.30

⬍.001

.25 39 (11.5) 299 (87.9) 26.0 (22.4-32.2)

5 (7.5) 62 (92.5) 21.1 (18.8-24.9)

⬍.001

NOTE. Data are presented as No. of patients (%), unless otherwise indicated. *Complete data missing for 1 subject. †Complete data missing for 6 subjects. Blakey et al. Third Molars and Periodontal Disease. J Oral Maxillofac Surg 2010.

visible groups was not significantly different (P ⬎ .05). Marginally more white subjects were represented in the not visible group (81.1%) than in the visible group (75.3%). Although the total number of enrolled nonwhite subjects was low, Asians predominated in the visible group compared with the not visible group: 7.6% versus 1.5%. Subjects in the visible group were significantly older, with a median age of 26 years (interquartile range [IQR], 22.4-32.2 years), as compared with the not visible group, with a median age of 21 years (IQR, 18.8-24.9 years) (P ⬍ .001). Subjects were highly educated; significantly more subjects with at least some college education were in the visible group (86.5%) than in the not visible group

(75.4%) (P ⬍ .001). This difference may reflect the mean age difference between the 2 groups. The rate of tobacco use was low in both groups and was only marginally higher in the visible group than in the not visible group: 11.5% versus 7.5%. Subjects in the visible group were significantly more likely to have at least 1 PD of 4 mm or greater on non–third molars than those in the not visible group (P ⬍ .001) (Table 2). In both groups, first/second molars were more likely to be affected than nonmolars: 59% versus 17% in the visible group and 35% versus 7% in the not visible group. Because the 2 groups differed significantly with respect to age, the data on non–third molar teeth were analyzed further

Table 2. PERCENTAGE OF SUBJECTS WITH AT LEAST 1 THIRD MOLAR VISIBLE (n ⴝ 342) (VISIBLE GROUP) AND THOSE WITH NO THIRD MOLARS VISIBLE (n ⴝ 69) (NOT VISIBLE GROUP) BY REGION OF MOUTH AND AT LEAST 1 NON–THIRD MOLAR PERIODONTAL PD OF 4 mm OR GREATER OR ALL NON–THIRD MOLAR PDs OF LESS THAN 4 mm

Subjects ⱖ1 Third molar visible (visible group) No third molars visible (not visible group)

All PDs ⬍4 mm ⱖ1 PD ⱖ4 mm First/Second Molar PD ⱖ4 mm Nonmolar PD ⱖ4 mm* [No. of Patients (%)] [No. of Patients (%)] [No. of Patients (%)] [No. of Patients (%)] 139 (41)

203 (59)

203 (59)

57 (17)

45 (65)

24 (35)

24 (35)

5 (7)

NOTE. Differences between groups are significant for all non–third molar teeth (P ⫽ .0002), first/second molars (P ⫽ .0002), and nonmolars (P ⫽ .046). *Subjects also had at least 1 PD of 4 mm or greater in first/second molars. Blakey et al. Third Molars and Periodontal Disease. J Oral Maxillofac Surg 2010.

328

THIRD MOLARS AND PERIODONTAL DISEASE

while we controlled for age. Subjects in the visible group were significantly more likely to have at least 1 non–third molar PD of 4 mm or greater than those in the not visible group, after adjustment for age less than 25 years and age of 25 years or greater (P ⬍ .001; odds ratio, 2.5 [95% confidence interval, 1.5-4.4]) (Table 3). On the basis of the total number of periodontal probing sites of 4 mm or greater per subject, severity of disease was low in these adolescent and young adult subjects; only beginning stages of periodontal inflammatory disease were detected (Table 4). However, the number of first/second molar PDs of 4 mm or greater tended to be higher for the visible group than for the not visible group, with a median of 1 (IQR, 0-4) versus 0 (IQR, 0-0), and for all teeth, with a median of 3 (IQR, 1-8) versus 0 (IQR, 0-0).

Discussion Our data suggest that healthy adolescents and young adults with visible asymptomatic third molars are more likely to have periodontal inflammatory disease detected on first/second molars as compared with those without visible third molars. The prevalence of periodontal inflammatory disease on first/

Table 3. PERCENTAGE OF SUBJECTS WITH AT LEAST 1 THIRD MOLAR VISIBLE (n ⴝ 342) (VISIBLE GROUP) AND THOSE WITH NO THIRD MOLARS VISIBLE (n ⴝ 69) (NOT VISIBLE GROUP) BY REGION OF MOUTH AND AT LEAST 1 NON–THIRD MOLAR PERIODONTAL PD OF 4 mm OR GREATER OR ALL NON–THIRD MOLAR PDs OF LESS THAN 4 mm ADJUSTED FOR AGE LESS THAN 25 YEARS OR AGE GREATER THAN OR EQUAL TO 25 YEARS

Subjects’ Non–Third Molar PD

Subjects Age ⬍25 y (n ⫽ 203) ⱖ1 Third molar visible (visible group) No third molars visible (not visible group) Age ⱖ25 y* (n ⫽ 206) ⱖ1 Third molar visible (visible group) No third molars visible (not visible group)

All PD ⬍4 mm [No. of Patients (%)]

ⱖ1 PD ⱖ4 mm [No. of Patients (%)]

66 (32)

85 (42)

36 (18)

16 (8)

72 (35)

117 (57)

9 (4)

8 (4)

NOTE. After adjustment for age (⬍25 years or ⱖ25 years), differences were significant for the visible and not visible groups (P ⫽ .0009; odds ratio, 2.5 [95% confidence interval, 1.5-4.4]). SAS Proc FREQ (SAS Institute, Cary, NC) was used to calculate differences by age. *Complete data missing for 2 subjects. Blakey et al. Third Molars and Periodontal Disease. J Oral Maxillofac Surg 2010.

Table 4. NUMBER OF PROBING SITES WITH PERIODONTAL PD OF 4 mm OR GREATER FOR SUBJECTS WITH AT LEAST 1 THIRD MOLAR VISIBLE (n ⴝ 342) (VISIBLE GROUP) AND SUBJECTS WITH NO THIRD MOLARS VISIBLE (n ⴝ 69) (NOT VISIBLE GROUP)

Number of Probing Sites With PDs ⱖ4 mm Per Subject

First or second molars Nonmolars All teeth

ⱖ1 Third Molar Visible (Visible Group) [Median (IQR)]

No Third Molars Visible (Not Visible Group) [Median (IQR)]

1 (0-4) 0 (0-0) 3 (1-8)

0 (0-1) 0 (0-0) 0 (0-0)

Blakey et al. Third Molars and Periodontal Disease. J Oral Maxillofac Surg 2010.

second molars in the visible group was almost twice that detected in the not visible group. In addition, the median number of periodontal probing sites of 4 mm or greater for all teeth was higher in the visible group, although overall periodontal inflammatory disease severity was low in both groups of study subjects. Although subjects in the visible group were, on average, significantly older than subjects in the not visible group, with a median of 26 years versus 21 years, the mean ages of both the visible and not visible groups were beyond the peak eruption period for third molars of 19.5 years suggested by Rantanen from a study of 2,000 university students.8 However, older subjects with visible third molars tended to be more likely to have periodontal inflammatory disease detected on first/second molars as compared with younger subjects: 57% versus 42%. Our findings are similar to those of Elter et al4 from young adult subjects aged 18 to 34 years in the population study NHANES III, although the study designs differ. Examinations for subjects in NHANES III were conducted in a nonclinical setting. Although the study design assessed the visible presence or absence of third molars, limited periodontal clinical data were collected and no radiographs were available. Only teeth in 1 randomly selected maxillary quadrant and 1 randomly selected mandibular quadrant were probed. Only 2 sites were probed, mesiobuccal and buccal, on up to 7 teeth in the quadrant. Periodontal probing data were not collected on third molars. With these limitations, Elter et al reported that a visible third molar doubled the odds of finding at least 1 PD of 5 mm or greater on the adjacent second molar. Moss et al9 included third molars in analyses of data from pregnant subjects. In a clinical study of almost 1,000 obstetric subjects aged less than 30 years, followed from the first trimester to after delivery, those with visible third molars were significantly more

329

BLAKEY ET AL

likely to have moderate or severe periodontal disease and less likely to be periodontally healthy at enrollment and postpartum than subjects with no visible third molars. PDs of at least 4 mm were detected significantly more often at interproximal molar sites in these subjects.10 Is it biologically plausible that visible third molars are reliable risk markers for possible periodontal inflammatory disease in adolescents and young adults? The combined impact of third molars often erupting incompletely at a later age after jaw growth is complete and of the anatomic location in the jaw of these teeth facilitates the accumulation of biofilm colonized by anaerobic periodontal pathogens in the molar regions of the jaw. Deeper PDs, of at least 4 mm, detected more often around third molars and particularly mandibular third molars have been associated with elevated subject level pathogen counts.11,12 Once established, the pathogenic bacteria are difficult to eradicate with mechanical debridement alone, and pathogens in molar sites may serve as potential reservoirs for pathogens colonizing more anterior in the mouth.13 Our findings suggest that the visible presence of asymptomatic third molars may be a risk marker for periodontal inflammatory disease on non–third molar teeth. Although not all visible asymptomatic third molars may be indicators of periodontal pathology, patients should be alerted that visible third molars and adjacent molars should be monitored over time for this possibility. However, our data should be used by clinicians with some caution. Although our study subjects are a diverse group of adolescents and young adults, they are not representative of the US population. Education levels were high, and Hispanic subjects were under-represented. Although tobacco use is a recognized risk factor for periodontal inflammatory disease, because few subjects with a PD ⱖ4 mm reported tobacco use, 26 of 227 (11.5%) subjects

were not stratified by tobacco use in analyses. Although we plan to continue to study our enrolled subjects over time, the topic of third molars and periodontal inflammatory disease in adolescents and young adults should be studied in other populations.

References 1. Ash MM, Costich ER, Hayward JR: A study of periodontal hazards of third molars. J Periodontol 33:209, 1962 2. Ash MM: Third molars as periodontal problems. Dent Clin North Am 8:51, 1964 3. Hugoson A, Kugelberg CF: The prevalence of third molars in a Swedish population, an epidemiological study. Community Dent Health 5:121, 1988 4. Elter JR, Cuomo C, Offenbacher S, et al: Third molars associated with periodontal pathology in the Third National Health and Nutrition Examination Survey. J Oral Maxillofac Surg 62: 440, 2004 5. Moss KL, Oh ES, Fisher E, et al: Third molars and periodontal pathology in middle-aged Americans. J Oral Maxillofac Surg 67:2592, 2009 6. Blakey GH, Hull D, Haug RH, et al: Changes in third molar and nonthird molar periodontal pathology over time. J Oral Maxillofac Surg 65:1577, 2007 7. White RP Jr, Phillips C, Hull DJ, et al: Risk markers for periodontal pathology over time in the third molar and non–third molar regions in young adults. J Oral Maxillofac Surg 66:749, 2008 8. Rantanen AV: The age of eruption of the third molar teeth. Acta Odontol Scand 25:64, 1967 (suppl 48) 9. Moss KL, Serlo AD, Offenbacher S, et al: The oral and systemic impact of third molar periodontal pathology. J Oral Maxillofac Surg 65:1739, 2007 10. Moss KL, Beck JD, Offenbacher S: Clinical risk factors associated with incidence and progression of periodontal conditions in pregnant women. J Clin Periodontal 32:492, 2005 11. Blakey GH, Jacks MT, Offenbacher S, et al: Progression of periodontal disease in the second/third molar region in patients with asymptomatic third molars. J Oral Maxillofac Surg 64:189, 2006 12. White RP Jr, Offenbacher S, Blakey GH, et al: Chronic oral inflammation and the progression of periodontal pathology in the third molar region. J Oral Maxillofac Surg 64:880, 2006 13. Moss KL, Serlo AD, Offenbacher S, et al: Third molars and the efficacy of mechanical debridement in reducing pathogen levels in pregnant subjects: A pilot study. J Oral Maxillofac Surg 66:1565, 2008