- Email: [email protected]
Third Molar Caries Experience in Middle-Aged and Older Americans: A Prevalence Study
J Oral Maxillofac Surg 68:634-640, 2010
Third Molar Caries Experience in Middle-Aged and Older Americans: A Prevalence Study Elda L. Fisher, DMD,* Kevin L. Moss,† Steven Offenbacher, DDS, PhD,‡ James D. Beck, PhD,§ and Raymond P. White, Jr, DDS, PhD储 Purpose: To assess the prevalence of third molar caries experience in a middle-aged and older
population and the relationship of these findings to caries experience in teeth more anterior in the mouth. Patients and Methods: Data from 6,550 Dental Atherosclerosis Risk in Communities study participants aged 52 to 74 years who underwent a clinical examination for coronal caries experience were available for these analyses. Subjects with visible third molars (N ⫽ 2,003) were divided based on third molar coronal caries experience on at least 1 third molar: no carious/decayed coronal surface (DS) or at least 1 coronal DS and no filled coronal surface (FS) or at least 1 coronal FS. Coronal DS and FS were also calculated for more anterior teeth. Covariates included ethnicity, gender, age, body mass index, education, income, smoking status, and diabetes diagnosis. Subject level outcomes for third molar and more anterior teeth were compared by descriptive statistics and 2 or t tests with statistical significance set at P less than .05. Multivariate modeling was performed to adjust outcome variables for covariates. Results: Third molar caries experience was detected in 77% of subjects and was significantly associated with caries experience in more anterior teeth and white race (P ⬍ .01). Caries experience was detected in only third molars in 1% of subjects, and 1% of subjects were caries free. Subjects with less education (20%) and lower income (19%) were significantly more likely to have DS detected compared with subjects with more education (6%) and higher income (5%) (P ⬍ .01). Conversely, subjects with more education (75%) and higher income (77%) were significantly more likely to have FS detected compared with subjects with less education (55%) and lower income (60%) (P ⬍ .01). Conclusions: Third molar coronal caries experience was significantly associated with caries experience in teeth more anterior in the mouth in this middle-aged and older population. © 2010 American Association of Oral and Maxillofacial Surgeons J Oral Maxillofac Surg 68:634-640, 2010 If patients have asymptomatic third molars, it is relevant for them to know the likelihood of retaining these teeth free of coronal caries for a lifetime. Third molar coronal caries is difficult to treat because of the anatomic location of these teeth. As a consequence, removal of affected third molars is often the most
practical treatment option. Because caries is an infectious disease affecting an individual patient, the treatment decision between third molar removal or third molar restoration should also depend on the patient’s overall caries experience.1-3 However, few epidemiologic studies reporting on the prevalence of caries
Received from the School of Dentistry, University of North Carolina, Chapel Hill, NC. *Resident, Department of Oral and Maxillofacial Surgery. †Research Specialist, Department of Dental Ecology. ‡OraPharma Distinguished Professor, Department of Periodontology. §Kenan Distinguished Professor and Associate Dean for Research, Department of Dental Ecology. 储Dalton L. McMichael Distinguished Professor, Department of Oral and Maxillofacial Surgery. Funding sources for these analyses were as follows: National
Institute of Dental and Craniofacial Research grant R01DE 11551, Oral and Maxillofacial Surgery Foundation, American Association of Oral and Maxillofacial Surgeons, and Dental Foundation of North Carolina. 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
0278-2391/10/6803-0023$36.00/0 doi:10.1016/j.joms.2009.10.003
634
635
FISHER ET AL
experience and carious lesions or restorations have included third molars.4 As a result, data are limited on the prevalence of third molar caries experience, as well as the relationship between the caries experience in third molars and caries experience in teeth more anterior in the mouth. Moss et al5 did report third molar caries experience in a population of senior adults with a mean age of 73 years. In that study 42% of 818 subjects had at least 1 visible third molar, and 28% of those subjects had evidence of coronal caries experience in a third molar. Shugars et al6 reported similar data on third molar caries experience in young adults with a mean age of 27 years who were enrolled with 4 asymptomatic third molars to study pathology developing in third molars over time. At enrollment, 29% of 211 subjects with at least 1 third molar at the occlusal plane had caries experience in third molars. These cross-sectional analyses were designed to expand our knowledge of third molar caries experience by assessing the prevalence of third molar coronal caries experience in a middle-aged and older population and the relationship of these findings to caries experience in teeth more anterior in the mouth.
Patients and Methods The Dental Atherosclerosis Risk in Communities (DARIC) study, an institutional review board–approved substudy of the Atherosclerosis Risk in Communities (ARIC) study, targeted 4 sites in the United States: Forsyth County, NC; Jackson, MS; suburbs of Minneapolis, MN; and Washington County, MD. The oral examination involved participants at ARIC visit 4 with data collected from 1996 –1998. These cross-sectional analyses included all dentate ARIC participants aged 52 to 74 years who underwent a clinical examination for coronal caries experience. The DARIC oral examination was conducted by trained, calibrated examiners. Clinical data on coronal caries were collected by visual-tactile examination according to the criteria described by Radike.7 Coronal caries experience— decayed and filled surfaces (DFS)— reflected existing restorations including crowns, as well as untreated carious lesions. No radiographs were available to complement the clinical examinations. Additional details of the DARIC study are available in an article by Beck et al.8 For our analyses, DARIC subjects with visible third molars were divided into non-exclusive groups based on subject level caries experience detected in at least 1 visible third molar: at least 1 carious/decayed coronal surface (DS) indicating untreated caries or no DS for each subject and at least 1 filled coronal surface (FS) indicating treated caries experience, restorations including crowns, or no FS for each subject. In addi-
tion, subject level coronal DS and FS were calculated for each remaining non–third molar tooth type. The main explanatory variable was the detection of subject level coronal caries experience (DS or FS) on visible third molars. The principal outcome variable was the subject level detection of coronal caries experience in at least 1 tooth type other than third molars. Potential confounders included study center/ ethnicity (African American or white), gender, age in years, body mass index (BMI) (weight [in kilograms]/ height [in square meters]), education (basic, 0-11 years; intermediate, 12-16 years; or advanced, ⱖ17 years), annual income level (⬍$25,000, $25,000-$50,000, or ⬎$50,000), smoking status (current heavy, current light, former heavy, former light, or never), and a medical history of diabetes mellitus. The subject level outcomes between those with and those without third molar caries experience were compared by descriptive statistics and bivariate analyses with 2 tests by use of SAS, version 9.2 (SAS Institute, Cary, NC). Statistical significance was set at P less than .05. The potential confounders— ethnicity, gender, age, BMI, education, income, smoking status, and diabetes—were added to adjusted models based on being associated with third molar caries experience. Multivariate modeling was performed by use of SAS Proc GLM to calculate least-squares means for outcome variables with adjustment for covariates.
Results Data on caries experience for these cross-sectional analyses were available for 6,550 of the 6,793 DARIC subjects (Fig 1). Caries data were incomplete for 243 subjects. No visible third molars were detected in 4,547 subjects (69%); at least 1 third molar was visible in 2,003 subjects (31%). Of those with at least 1 visible third molar (ie, the subjects studied in these analyses), 954 subjects (48%) had only 1 third molar detected; at least 3 third molars were detected in 481 subjects (24%). African-American subjects were more likely to have at least 1 visible third molar than were white subjects (40% vs 29%) (data not shown). Of those patients with visible third molars, caries experience was detected in at least 1 third molar coronal surface in 1,550 subjects (77%); third molars were free of caries experience (ie, untreated caries or restorations) in 453 subjects (23%) (Table 1). In teeth more anterior to third molars, caries experience was detected in at least 1 tooth in 1,960 subjects (98%); 43 subjects had no caries experience in these teeth (2%) (Table 1). Coronal third molar caries experience was significantly associated with coronal caries experience in more anterior teeth for 1,530 subjects (76%) (P ⬍ .01). Caries experience was exclusive to third
636
FIGURE 1. Flow diagram for DARIC subjects and analyses of third molar caries experience. Asterisk, Caries experience data missing for 243 subjects. Dagger, Remaining teeth other than third molars. Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
molars in 20 subjects (1%), and 23 subjects (1%) were caries free. On the basis of the design of the overall ARIC study, 372 subjects in Jackson, MS, and 48 subjects in North Carolina (420 of 2,003 subjects [21%]) were African American; the remaining 1,583 subjects (79%) were white (Table 2). African Americans were significantly more likely to have third molars without caries experience (ie, no DFS) compared with white subjects (34% vs 19%, P ⬍ .01). Subjects with caries-free third molars were also significantly less educated (less than completion of grade 12) (P ⫽ .04), significantly more likely to be former heavy or current tobacco users (P ⫽ .01), and significantly more likely to have a diabetes diagnosis (P ⫽ .02). Gender, age, BMI, and annual income were not significantly associated with third molar coronal caries experience (P ⱖ .05). Third molar DFS was significantly associated with DFS for each other tooth type and with subjects’ overall caries experience, which included all teeth other than third molars (P ⬍ .01) (Table 3). Significantly fewer subjects overall had untreated third molar caries (DS) (153 of 2,003 [8%]) than treated caries (FS) (1,441 of 2,003 [73%]) (P ⬍ .01) (Table 4). No significant differences for third molar DS or FS computed for all subjects compared by age or gender were detected. Significantly more AfricanAmerican subjects (25%) had DS detected compared with white subjects (3%) (P ⬍ .01) (Table 4). Conversely, significantly more white subjects (80%) had
THIRD MOLAR CARIES
FS detected compared with African-American subjects (46%) (P ⬍ .01). Subjects with less education (less than completion of grade 12) (20%) and lower income (⬍$25,000 annually) (19%) were significantly more likely to have DS detected compared with subjects with more education (6%) and higher income (5%) (both P ⬍ .01). Conversely, subjects with more education (75%) and higher income (77%) were significantly more likely to have FS detected compared with subjects with less education (55%) and lower income (60%) (both P ⬍ .01). Subjects with a higher BMI were significantly more likely to have DS and less likely to have FS compared with subjects with lower BMI (P ⬍ .01). Significantly more subjects with diabetes mellitus had DS (16%) than subjects without a diabetes history (6%) (P ⬍ .01). Conversely, significantly more subjects without diabetes had FS (75%) compared with subjects with diabetes mellitus (61%) (P ⬍ .01). For the DARIC population overall, subjects reporting never using tobacco or former light tobacco use were significantly more likely to have third molar DS or FS (P ⬍ .01) (data not shown). Subjects who reported never using tobacco or former light tobacco use also were significantly more likely to have FS (75%) compared with subjects with former heavy or current tobacco use (70%) (P ⫽ .01) (Table 4). No significant differences were detected for DS by tobacco use. After adjustment for covariates, third molar DS and FS were significantly associated with detection of DS and FS in all other tooth types (P ⬍ .01) (data not shown). If DS was detected in third molars, mean DS was significantly greater in each tooth type more anterior in the mouth (P ⬍ .01) (Table 5). If FS was detected in third molars, mean FS was significantly greater in first/second molars and second premolars (both P ⬍ .01) and progressively lesser in anterior teeth (ie, canines and incisors) (all P ⫽ .02).
Table 1. DARIC SUBJECTS WITH AT LEAST 1 VISIBLE THIRD MOLAR (N ⴝ 2,003) DISPLAYED BY THIRD MOLAR AND NON–THIRD MOLAR CARIES EXPERIENCE (DFS)
Teeth Other Than Third Molars No DFS ⱖ1 DFS Total
Third Molar [n (%)] No DFS
ⱖ1 DFS
Total
23 (1) 430 (21) 453 (23)
20 (1) 1,530 (76) 1,550 (77)
43 (2) 1,960 (98) 2,003 (100)
NOTE. At least 1 third molar DFS was significantly associated with DFS in at least 1 other remaining tooth (P ⬍ .01). Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
637
FISHER ET AL
Table 2. DEMOGRAPHIC AND SOCIAL CHARACTERISTICS OF DARIC SUBJECTS WITH THIRD MOLAR CARIES EXPERIENCE (DFS) IN AT LEAST 1 THIRD MOLAR OR NONE PER SUBJECT (N ⴝ 2,003)
No Third Molar With DFS Ethnicity/center [n (%)] African American Jackson, MS Forsyth County, NC Total White Forsyth County, NC Washington County, MD Suburbs of Minneapolis, MN Total Gender [n (%)] Male Female Age [mean (SE)] (yr) BMI age [mean (SE)] Education [n (%)] Basic (0-ⱕ11 yr) Intermediate (12-16 yr) Advanced (ⱖ17 yr) Total Annual income [n (%)] Low (⬍$25,000) Medium ($25,000-$50,000) High (⬎$50,000) Total Tobacco use [n (%)] Never Former light Total Former heavy Current light Current heavy Total Diabetes diagnosis [n (%)] Yes No Total
ⱖ1 Third Molar With DFS
P Value
129 (34.7) 15 (31.3) 144 (34.3)
243 (65.3) 33 (68.8) 276 (65.7)
106 (22.3) 85 (17.7) 111 (18.3) 302 (19.3)
370 (77.7) 395 (82.3) 496 (81.7) 1,261 (80.7)
⬍.01
247 (21.8) 206 (23.7) 61.3 (5.5) 28.8 (5.0)
888 (78.2) 662 (76.3) 61.6 (5.5) 28.2 (5.1)
.30 .28 .05
68 (28.8) 180 (20.9) 205 (22.7) 385 (21.8)
168 (71.2) 680 (79.1) 700 (77.4) 1,380 (78.2)
.04
97 (24.7) 161 (23.0) 168 (20.4) 329 (21.6)
296 (75.3) 540 (77.0) 655 (79.6) 1,195 (78.4)
198 (20.7) 104 (21.5) 302 (21.0) 65 (22.3) 18 (42.9) 44 (28.6) 127 (26.0)
757 (79.3) 379 (78.5) 1,136 (79.0) 226 (77.7) 24 (57.1) 110 (71.4) 360 (74.0)
.01
76 (28.0) 368 (21.5) 444 (22.3)
195 (72.0) 1,345 (78.5) 1,540 (77.7)
.02
.21
NOTE. African Americans were significantly more likely to have no third molar DFS compared with white subjects (P ⬍ .01). Subjects who were less well educated (basic [less than a high school education]) were significantly more likely to have no third molar DFS compared with those more highly educated (P ⫽ .04). Subjects who were former heavy or current tobacco users were significantly more likely to have no third molar DFS compared with those with lesser tobacco use (P ⫽ .01). Subjects with a diabetes diagnosis were significantly more likely to have no third molar DFS compared with those without a diabetes diagnosis (P ⫽ .02). Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
Discussion The most clinically relevant findings from these cross-sectional analyses of data from middle and olderaged subjects were as follows. Very few DARIC subjects, only 1% of the 2,003 DARIC subjects with visible third molars, were free of caries experience (ie, no DFS in any teeth including third molars). Seventyseven percent of subjects had coronal caries experience detected in at least 1 third molar. Caries experience was not unique to third molars: only 1% of subjects had DFS exclusively in third molars. Therefore almost all the subjects with third molar DFS also
had caries experience in more anterior teeth. Although fewer subjects had third molar DS overall compared with third molar FS, African Americans were significantly more likely to have third molar DS and less likely to have third molar FS compared with white subjects. Lower education level, lower income, or greater BMI were significantly associated with a greater number of subjects having third molar DS detected. Conversely, higher education level, higher income, and lower BMI were significantly associated with a greater number of subjects having third molar FS detected.
638
THIRD MOLAR CARIES
Table 3. MEAN (SD) DFS FOR DARIC SUBJECTS BY TOOTH TYPE COMPARED BY CORONAL CARIES EXPERIENCE IN AT LEAST 1 THIRD MOLAR (DFS) OR NONE (N ⴝ 2,003)
Mean DFS (SD) No Third Molar With DFS
ⱖ1 Third Molar With DFS
All remaining non–third molar teeth 12.1 (10.2) 19.4 (11.7) Second molars 2.6 (2.6) 4.8 (3.3) First molars 2.7 (3.1) 4.4 (3.8) Second premolars 2.6 (2.9) 4.1 (3.2) First premolars 2.2 (2.4) 3.3 (2.8) Canines 0.9 (1.6) 1.2 (1.8) Lateral incisors 0.9 (1.5) 1.2 (1.7) Central incisors 1.0 (1.6) 1.3 (1.8)
P Value ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001 .004 .0007 ⬍.0001
Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
How do the DARIC data on subjects retaining third molars compare with the few population data on the prevalence of visible third molars in the existing literature? Hugoson and Kugelberg9 have reported the only population data on the prevalence of third molars. The number of DARIC subjects with at least 1 visible third molar was lower than the number reported by Hugoson and Kugelberg for the cohort aged 60 years (31% vs 44%). However, in DARIC subjects with visible third molars we report a greater number of third molars compared with Hugoson and Kugelberg: 48% with only 1 third molar and 24% with at least 3 third molars compared with 21% and 12%, respectively. Not all third molars present in Hugoson and Kugelberg’s reported data were visible; some were covered, detected only by radiographs, and unlikely to have caries experience, further increasing the discrepancy in number of third molars and the prevalence of third molar caries experience in DARIC subjects. The 19% of subjects with visible third molars in the cohort aged 70 years reported by Hugoson and Kugelberg9 and the 31% of DARIC subjects were fewer than the 42% of 818 senior subjects in the analyses of Moss et al5 with visible third molars, who had a mean age of 73 years. However, we report a much higher prevalence of third molar caries experience in DARIC subjects than did Moss et al from senior subjects aged a decade older enrolled in the Piedmont 65⫹ population study (77% vs 28%). Caries experience was also more prevalent in teeth anterior to third molars in DARIC subjects with a visible third molar compared with the Piedmont 65⫹ subjects (98% vs 50%).5 In cross-sectional analyses of data from Piedmont 65⫹ and DARIC, third molar caries experience was significantly associated with caries experience in teeth more anterior in the mouth. This relationship alone may account for these differences.
Our data indicating that third molar caries experience was not exclusive to third molars is consistent with other reports. In the younger population of Shugars et al,6 with a mean age of 27 years, only 1% of 211 subjects had third molar caries experience without other teeth being affected. Moss et al5 reported that 8% of Piedmont 65⫹ subjects had caries experience exclusive to third molars; other retained teeth were not affected. Although these outcomes may be surprising to some clinicians, the findings are plausible biologically. Caries experience requires the colonization of specific, acidforming bacteria in biofilm attached to a non-sheddable surface of susceptible teeth. If the requisite micro-organisms are present in the biofilm on 1 tooth, all visible teeth in the mouth are at risk for a lifetime. Thus, for individuals with caries experience in teeth other than third molars earlier in life, the risk for third molar caries experience may be greater because third molars erupt over time. Conversely, if patients have no caries experience, third molar caries seems less likely. The prevalence of third molar caries experience may differ by ethnic group. A greater number of African-American DARIC subjects had at least 1 visible third molar than did white subjects (40% vs 29%), potentially increasing the risk for the prevalence of third molar caries experience in African Americans. However, we report that caries-free third molars (ie, no DS or FS) were significantly more prevalent in African Americans than white subjects. Moss et al10 reported similar results from the Piedmont 65⫹ subjects, an older population with a mean age of 73 years; white subjects were significantly more likely to have caries experience than African Americans (odds ratio, 2.5 [95% confidence interval, 1.4-4.3]). In our analyses, African Americans had a greater prevalence of third molar DS than white subjects. Drake et al11 reported similar findings from third molars and other remaining teeth in Piedmont 65⫹ subjects. Because half of the subjects studied in Piedmont 65⫹ were lifelong residents of a rural environment, Drake et al suggested that the differences were related to a lack of access to care. Our findings significantly associating lower income and lower education level with greater third molar DS could also be attributed to the same lack of access to care. The significant positive relationship between a history of diabetes and DS may be related to poorer overall health status, often complicated by less access to health care at some period in the affected individual’s life. The significant positive relationships we report between no use of tobacco or light former tobacco use or no diabetes diagnosis and having third molar FS was also reported by Moss et al10 from the Piedmont 65⫹ population; no tobacco use significantly increased the
639
FISHER ET AL
Table 4. DEMOGRAPHIC AND SOCIAL CHARACTERISTICS OF DARIC SUBJECTS WITH THIRD MOLAR CARIES EXPERIENCE (DS OR FS) IN AT LEAST 1 THIRD MOLAR OR NONE PER SUBJECT (N ⴝ 2,003)
No Third Molar ⱖ1 Third Molar No Third Molar ⱖ1 Third Molar With DS With DS P Value With FS With FS P Value Ethnicity/center [n (%)] African American Jackson, MS Forsyth County, NC Total White Forsyth County, NC Washington County, MD Suburbs of Minneapolis, MN Total Gender [n (%)] Male Female
277 (74.5) 40 (83.3) 317 (75.5)
95 (25.5) 8 (16.7) 103 (24.5)
206 (55.4) 21 (43.8) 227 (54.0)
166 (44.6) 27 (56.3) 193 (46.0)
470 (98.7) 448 (93.3) 595 (98.0) 1,513 (96.8)
6 (1.3) 32 (6.7) 12 (2.0) 50 (3.2)
⬍.01
107 (22.5) 96 (20.0) 112 (18.5) 315 (10.2)
369 (77.5) 384 (80.0) 495 (81.6) 1,248 (79.8)
⬍.01
1,039 (91.5) 809 (93.2)
96 (8.5) 59 (6.8)
.17
302 (26.6) 248 (28.6)
833 (73.4) 620 (71.4)
.33
Age [mean (SE)] (yr)
61.6 (5.5)
61.1 (5.1)
.35
61.3 (5.4)
61.6 (5.5)
.25
BMI [mean (SE)]
28.2 (4.9)
30.3 (6.3)
⬍.01
29.1 (5.2)
28.1 (5.0)
⬍.01
⬍.01
107 (45.3) 218 (5.4) 225 (24.9) 443 (25.0)
129 (54.7) 642 (74.7) 680 (75.1) 1,322 (75.0)
⬍.01
⬍.01
159 (40.5) 182 (26.0) 175 (21.3) 357 (23.4)
234 (59.5) 519 (74.0) 648 (78.7) 1,167 (76.6)
⬍.01
.08
247 (25.9) 123 (25.5) 370 (25.7) 71 (24.4) 23 (54.8) 52 (33.8) 146 (30.0)
708 (74.1) 360 (74.5) 1,068 (74.3) 220 (75.6) 19 (45.2) 102 (66.2) 341 (70.0)
.01
⬍.01
106 (39.1) 432 (25.2) 538 (27.1)
165 (60.9) 1,281 (74.8) 1,446 (72.9)
⬍.01
Education [n (%)] Basic (0-11 yr) Intermediate (12-16 yr) Advanced (ⱖ17 yr) Total Annual income [n (%)] Low (⬍$25,000) Medium ($25,000-$50,000) High (⬎$50,000) Total Tobacco use [n (%)] Never Former light Total Former heavy Current light Current heavy Total Diabetes diagnosis [n (%)] Yes No Total
189 (80.1) 801 (93.1) 856 (94.6) 1,657 (93.9)
47 (19.9) 59 (6.9) 49 (5.4) 108 (6.1)
319 (81.2) 660 (94.2) 792 (96.2) 1,452 (95.3)
74 (18.8) 41 (5.9) 31 (3.8) 72 (4.7)
880 (92.2) 445 (92.1) 1,325 (92.1) 279 (95.9) 36 (85.7) 140 (90.9) 455 (93.4)
75 (7.9) 38 (7.9) 113 (7.9) 12 (4.1) 6 (14.3) 14 (9.1) 32 (6.6)
229 (84.5) 1,604 (93.6) 1,833 (92.4)
42 (15.5) 109 (6.4) 151 (7.6)
Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
odds of detecting third molar caries experience (odds ratio, 2.3 [95% confidence interval, 1.1-4.4]). Although we cannot document a relationship between better overall health and FS, it is plausible that individuals who never or rarely use tobacco or do not have diabetes are in better health and may seek health care more frequently, including restorations for caries once it is detected. The association should be studied further in other populations. The findings we report from ARIC subjects have limits and may not be applicable to all middle-aged and older individuals. Although our study population numbers over 6,500 subjects, only 4 communities in the United States were represented. A greater percentage of African-American subjects was included
compared with the US population. Subjects who were not white or African American were excluded from analyses because few subjects were enrolled. Thus Asian and Latino subjects in the United States were not studied. Although few unerupted third molars would be affected by caries experience, the number of third molars in the ARIC subjects may be understated because radiographs to detect impacted teeth were not available to examiners. The few population studies documenting the presence or absence of third molars, including the data from ARIC, do not report whether missing third molars were removed to treat existing disease or to prevent future pathology, nor do the studies document the relationship between third molars and the number of other teeth that may
640
THIRD MOLAR CARIES
Table 5. ADJUSTED MEAN (SE) DS AND FS FOR DARIC SUBJECTS BY TOOTH TYPE COMPARED BY THIRD MOLAR CORONAL CARIES EXPERIENCE (DS OR FS) IN AT LEAST 1 THIRD MOLAR OR NONE (N ⴝ 2,003)
Adjusted Mean (SE)*
Adjusted Mean (SE)*
Tooth Type
No Third Molar With DS
ⱖ1 Third Molar With DS
P Value
No Third Molar With FS
ⱖ1 Third Molar With FS
P Value
Second molars First molars Second premolars First premolars Canines Lateral incisors Central incisors
0.08 (0.01) 0.09 (0.01) 0.09 (0.01) 0.05 (0.01) 0.03 (0.01) 0.04 (0.01) 0.04 (0.01)
0.63 (0.05) 0.65 (0.05) 0.44 (0.05) 0.25 (0.03) 0.22 (0.02) 0.31 (0.03) 0.28 (0.03)
⬍.01 ⬍.01 ⬍.01 ⬍.01 ⬍.01 ⬍.01 ⬍.01
2.69 (0.14) 2.84 (0.18) 2.65 (0.14) 2.28 (0.12) 0.94 (0.08) 0.97 (0.07) 0.98 (0.08)
4.72 (0.08) 4.40 (0.10) 4.09 (0.08) 3.25 (0.07) 1.16 (0.05) 1.16 (0.04) 1.21 (0.05)
⬍.01 ⬍.01 ⬍.01 ⬍.01 .02 .02 .02
*Adjusted for ethnicity/center, age, gender, tobacco use, education level, income level, BMI, and history of diabetes mellitus. Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
be present or missing. As a consequence, our data on prevalence of caries experience in retained third molars and the association with caries experience overall may be understated or overstated. However, with these limitations, the data we report do add to our understanding of caries experience, a chronic infectious disease, and the relationship of third molar caries experience in retained third molars to caries experience in other more anterior teeth. As other populations are studied for caries experience, a clinical caries assessment of third molars should be included. Acknowledgments The ARIC study is being carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022. The authors thank the staff and participants of the ARIC study for their important contributions. This manuscript was approved by the ARIC Coordinating Committee.
References 1. Murdoch-Kinch CA, McLean ME: Minimally invasive dentistry. J Am Dent Assoc 134:87, 2003
2. Powell LV: Caries risk assessment: Relevance to the practitioner. J Am Dent Assoc 129:349, 1998 3. Powell LV: Caries prediction: A review of the literature. Community Dent Oral Epidemiol 26:361, 1998 4. Brown LJ, Wall TP, Lazar V: Trends in caries among adults 18 to 45 years old. J Am Dent Assoc 133:827, 2002 5. Moss KL, Beck JD, Mauriello SM, et al: Third molar periodontal disease and caries in senior adults. J Oral Maxillofac Surg 65:103, 2007 6. Shugars DA, Elter JR, Jacks TM, et al: Incidence of occlusal dental caries in asymptomatic third molars. J Oral Maxillofac Surg 63:341, 2005 7. Radike A: Criteria for the diagnosis of dental caries, in Proceedings of the Conference on Clinical Testing of Cariostatic Agents. Chicago, American Dental Association, 1968, p 88 8. Beck JD, Elter JR, Heiss G, et al: Relationship of periodontal disease to carotid artery intima-media wall thickness: The atherosclerosis risk in communities (ARIC) study. Arterioscler Thromb Vasc Biol 21:1816, 2001 9. Hugoson A, Kugelberg CF: The prevalence of third molars in a Swedish population. An epidemiological study. Community Dent Health 5:121, 1988 10. Moss KL, Beck JD, Mauriello SM, et al: Risk indicators for third molar caries and periodontal disease in senior adults. J Oral Maxillofac Surg 65:958, 2007 11. Drake CW, Beck JD, Lawrence HP, et al: Three-year coronal caries incidence and risk factors in North Carolina elderly. Caries Res 31:1, 1997
Third Molar Caries Experience in Middle-Aged and Older Americans: A Prevalence Study Elda L. Fisher, DMD,* Kevin L. Moss,† Steven Offenbacher, DDS, PhD,‡ James D. Beck, PhD,§ and Raymond P. White, Jr, DDS, PhD储 Purpose: To assess the prevalence of third molar caries experience in a middle-aged and older
population and the relationship of these findings to caries experience in teeth more anterior in the mouth. Patients and Methods: Data from 6,550 Dental Atherosclerosis Risk in Communities study participants aged 52 to 74 years who underwent a clinical examination for coronal caries experience were available for these analyses. Subjects with visible third molars (N ⫽ 2,003) were divided based on third molar coronal caries experience on at least 1 third molar: no carious/decayed coronal surface (DS) or at least 1 coronal DS and no filled coronal surface (FS) or at least 1 coronal FS. Coronal DS and FS were also calculated for more anterior teeth. Covariates included ethnicity, gender, age, body mass index, education, income, smoking status, and diabetes diagnosis. Subject level outcomes for third molar and more anterior teeth were compared by descriptive statistics and 2 or t tests with statistical significance set at P less than .05. Multivariate modeling was performed to adjust outcome variables for covariates. Results: Third molar caries experience was detected in 77% of subjects and was significantly associated with caries experience in more anterior teeth and white race (P ⬍ .01). Caries experience was detected in only third molars in 1% of subjects, and 1% of subjects were caries free. Subjects with less education (20%) and lower income (19%) were significantly more likely to have DS detected compared with subjects with more education (6%) and higher income (5%) (P ⬍ .01). Conversely, subjects with more education (75%) and higher income (77%) were significantly more likely to have FS detected compared with subjects with less education (55%) and lower income (60%) (P ⬍ .01). Conclusions: Third molar coronal caries experience was significantly associated with caries experience in teeth more anterior in the mouth in this middle-aged and older population. © 2010 American Association of Oral and Maxillofacial Surgeons J Oral Maxillofac Surg 68:634-640, 2010 If patients have asymptomatic third molars, it is relevant for them to know the likelihood of retaining these teeth free of coronal caries for a lifetime. Third molar coronal caries is difficult to treat because of the anatomic location of these teeth. As a consequence, removal of affected third molars is often the most
practical treatment option. Because caries is an infectious disease affecting an individual patient, the treatment decision between third molar removal or third molar restoration should also depend on the patient’s overall caries experience.1-3 However, few epidemiologic studies reporting on the prevalence of caries
Received from the School of Dentistry, University of North Carolina, Chapel Hill, NC. *Resident, Department of Oral and Maxillofacial Surgery. †Research Specialist, Department of Dental Ecology. ‡OraPharma Distinguished Professor, Department of Periodontology. §Kenan Distinguished Professor and Associate Dean for Research, Department of Dental Ecology. 储Dalton L. McMichael Distinguished Professor, Department of Oral and Maxillofacial Surgery. Funding sources for these analyses were as follows: National
Institute of Dental and Craniofacial Research grant R01DE 11551, Oral and Maxillofacial Surgery Foundation, American Association of Oral and Maxillofacial Surgeons, and Dental Foundation of North Carolina. 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
0278-2391/10/6803-0023$36.00/0 doi:10.1016/j.joms.2009.10.003
634
635
FISHER ET AL
experience and carious lesions or restorations have included third molars.4 As a result, data are limited on the prevalence of third molar caries experience, as well as the relationship between the caries experience in third molars and caries experience in teeth more anterior in the mouth. Moss et al5 did report third molar caries experience in a population of senior adults with a mean age of 73 years. In that study 42% of 818 subjects had at least 1 visible third molar, and 28% of those subjects had evidence of coronal caries experience in a third molar. Shugars et al6 reported similar data on third molar caries experience in young adults with a mean age of 27 years who were enrolled with 4 asymptomatic third molars to study pathology developing in third molars over time. At enrollment, 29% of 211 subjects with at least 1 third molar at the occlusal plane had caries experience in third molars. These cross-sectional analyses were designed to expand our knowledge of third molar caries experience by assessing the prevalence of third molar coronal caries experience in a middle-aged and older population and the relationship of these findings to caries experience in teeth more anterior in the mouth.
Patients and Methods The Dental Atherosclerosis Risk in Communities (DARIC) study, an institutional review board–approved substudy of the Atherosclerosis Risk in Communities (ARIC) study, targeted 4 sites in the United States: Forsyth County, NC; Jackson, MS; suburbs of Minneapolis, MN; and Washington County, MD. The oral examination involved participants at ARIC visit 4 with data collected from 1996 –1998. These cross-sectional analyses included all dentate ARIC participants aged 52 to 74 years who underwent a clinical examination for coronal caries experience. The DARIC oral examination was conducted by trained, calibrated examiners. Clinical data on coronal caries were collected by visual-tactile examination according to the criteria described by Radike.7 Coronal caries experience— decayed and filled surfaces (DFS)— reflected existing restorations including crowns, as well as untreated carious lesions. No radiographs were available to complement the clinical examinations. Additional details of the DARIC study are available in an article by Beck et al.8 For our analyses, DARIC subjects with visible third molars were divided into non-exclusive groups based on subject level caries experience detected in at least 1 visible third molar: at least 1 carious/decayed coronal surface (DS) indicating untreated caries or no DS for each subject and at least 1 filled coronal surface (FS) indicating treated caries experience, restorations including crowns, or no FS for each subject. In addi-
tion, subject level coronal DS and FS were calculated for each remaining non–third molar tooth type. The main explanatory variable was the detection of subject level coronal caries experience (DS or FS) on visible third molars. The principal outcome variable was the subject level detection of coronal caries experience in at least 1 tooth type other than third molars. Potential confounders included study center/ ethnicity (African American or white), gender, age in years, body mass index (BMI) (weight [in kilograms]/ height [in square meters]), education (basic, 0-11 years; intermediate, 12-16 years; or advanced, ⱖ17 years), annual income level (⬍$25,000, $25,000-$50,000, or ⬎$50,000), smoking status (current heavy, current light, former heavy, former light, or never), and a medical history of diabetes mellitus. The subject level outcomes between those with and those without third molar caries experience were compared by descriptive statistics and bivariate analyses with 2 tests by use of SAS, version 9.2 (SAS Institute, Cary, NC). Statistical significance was set at P less than .05. The potential confounders— ethnicity, gender, age, BMI, education, income, smoking status, and diabetes—were added to adjusted models based on being associated with third molar caries experience. Multivariate modeling was performed by use of SAS Proc GLM to calculate least-squares means for outcome variables with adjustment for covariates.
Results Data on caries experience for these cross-sectional analyses were available for 6,550 of the 6,793 DARIC subjects (Fig 1). Caries data were incomplete for 243 subjects. No visible third molars were detected in 4,547 subjects (69%); at least 1 third molar was visible in 2,003 subjects (31%). Of those with at least 1 visible third molar (ie, the subjects studied in these analyses), 954 subjects (48%) had only 1 third molar detected; at least 3 third molars were detected in 481 subjects (24%). African-American subjects were more likely to have at least 1 visible third molar than were white subjects (40% vs 29%) (data not shown). Of those patients with visible third molars, caries experience was detected in at least 1 third molar coronal surface in 1,550 subjects (77%); third molars were free of caries experience (ie, untreated caries or restorations) in 453 subjects (23%) (Table 1). In teeth more anterior to third molars, caries experience was detected in at least 1 tooth in 1,960 subjects (98%); 43 subjects had no caries experience in these teeth (2%) (Table 1). Coronal third molar caries experience was significantly associated with coronal caries experience in more anterior teeth for 1,530 subjects (76%) (P ⬍ .01). Caries experience was exclusive to third
636
FIGURE 1. Flow diagram for DARIC subjects and analyses of third molar caries experience. Asterisk, Caries experience data missing for 243 subjects. Dagger, Remaining teeth other than third molars. Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
molars in 20 subjects (1%), and 23 subjects (1%) were caries free. On the basis of the design of the overall ARIC study, 372 subjects in Jackson, MS, and 48 subjects in North Carolina (420 of 2,003 subjects [21%]) were African American; the remaining 1,583 subjects (79%) were white (Table 2). African Americans were significantly more likely to have third molars without caries experience (ie, no DFS) compared with white subjects (34% vs 19%, P ⬍ .01). Subjects with caries-free third molars were also significantly less educated (less than completion of grade 12) (P ⫽ .04), significantly more likely to be former heavy or current tobacco users (P ⫽ .01), and significantly more likely to have a diabetes diagnosis (P ⫽ .02). Gender, age, BMI, and annual income were not significantly associated with third molar coronal caries experience (P ⱖ .05). Third molar DFS was significantly associated with DFS for each other tooth type and with subjects’ overall caries experience, which included all teeth other than third molars (P ⬍ .01) (Table 3). Significantly fewer subjects overall had untreated third molar caries (DS) (153 of 2,003 [8%]) than treated caries (FS) (1,441 of 2,003 [73%]) (P ⬍ .01) (Table 4). No significant differences for third molar DS or FS computed for all subjects compared by age or gender were detected. Significantly more AfricanAmerican subjects (25%) had DS detected compared with white subjects (3%) (P ⬍ .01) (Table 4). Conversely, significantly more white subjects (80%) had
THIRD MOLAR CARIES
FS detected compared with African-American subjects (46%) (P ⬍ .01). Subjects with less education (less than completion of grade 12) (20%) and lower income (⬍$25,000 annually) (19%) were significantly more likely to have DS detected compared with subjects with more education (6%) and higher income (5%) (both P ⬍ .01). Conversely, subjects with more education (75%) and higher income (77%) were significantly more likely to have FS detected compared with subjects with less education (55%) and lower income (60%) (both P ⬍ .01). Subjects with a higher BMI were significantly more likely to have DS and less likely to have FS compared with subjects with lower BMI (P ⬍ .01). Significantly more subjects with diabetes mellitus had DS (16%) than subjects without a diabetes history (6%) (P ⬍ .01). Conversely, significantly more subjects without diabetes had FS (75%) compared with subjects with diabetes mellitus (61%) (P ⬍ .01). For the DARIC population overall, subjects reporting never using tobacco or former light tobacco use were significantly more likely to have third molar DS or FS (P ⬍ .01) (data not shown). Subjects who reported never using tobacco or former light tobacco use also were significantly more likely to have FS (75%) compared with subjects with former heavy or current tobacco use (70%) (P ⫽ .01) (Table 4). No significant differences were detected for DS by tobacco use. After adjustment for covariates, third molar DS and FS were significantly associated with detection of DS and FS in all other tooth types (P ⬍ .01) (data not shown). If DS was detected in third molars, mean DS was significantly greater in each tooth type more anterior in the mouth (P ⬍ .01) (Table 5). If FS was detected in third molars, mean FS was significantly greater in first/second molars and second premolars (both P ⬍ .01) and progressively lesser in anterior teeth (ie, canines and incisors) (all P ⫽ .02).
Table 1. DARIC SUBJECTS WITH AT LEAST 1 VISIBLE THIRD MOLAR (N ⴝ 2,003) DISPLAYED BY THIRD MOLAR AND NON–THIRD MOLAR CARIES EXPERIENCE (DFS)
Teeth Other Than Third Molars No DFS ⱖ1 DFS Total
Third Molar [n (%)] No DFS
ⱖ1 DFS
Total
23 (1) 430 (21) 453 (23)
20 (1) 1,530 (76) 1,550 (77)
43 (2) 1,960 (98) 2,003 (100)
NOTE. At least 1 third molar DFS was significantly associated with DFS in at least 1 other remaining tooth (P ⬍ .01). Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
637
FISHER ET AL
Table 2. DEMOGRAPHIC AND SOCIAL CHARACTERISTICS OF DARIC SUBJECTS WITH THIRD MOLAR CARIES EXPERIENCE (DFS) IN AT LEAST 1 THIRD MOLAR OR NONE PER SUBJECT (N ⴝ 2,003)
No Third Molar With DFS Ethnicity/center [n (%)] African American Jackson, MS Forsyth County, NC Total White Forsyth County, NC Washington County, MD Suburbs of Minneapolis, MN Total Gender [n (%)] Male Female Age [mean (SE)] (yr) BMI age [mean (SE)] Education [n (%)] Basic (0-ⱕ11 yr) Intermediate (12-16 yr) Advanced (ⱖ17 yr) Total Annual income [n (%)] Low (⬍$25,000) Medium ($25,000-$50,000) High (⬎$50,000) Total Tobacco use [n (%)] Never Former light Total Former heavy Current light Current heavy Total Diabetes diagnosis [n (%)] Yes No Total
ⱖ1 Third Molar With DFS
P Value
129 (34.7) 15 (31.3) 144 (34.3)
243 (65.3) 33 (68.8) 276 (65.7)
106 (22.3) 85 (17.7) 111 (18.3) 302 (19.3)
370 (77.7) 395 (82.3) 496 (81.7) 1,261 (80.7)
⬍.01
247 (21.8) 206 (23.7) 61.3 (5.5) 28.8 (5.0)
888 (78.2) 662 (76.3) 61.6 (5.5) 28.2 (5.1)
.30 .28 .05
68 (28.8) 180 (20.9) 205 (22.7) 385 (21.8)
168 (71.2) 680 (79.1) 700 (77.4) 1,380 (78.2)
.04
97 (24.7) 161 (23.0) 168 (20.4) 329 (21.6)
296 (75.3) 540 (77.0) 655 (79.6) 1,195 (78.4)
198 (20.7) 104 (21.5) 302 (21.0) 65 (22.3) 18 (42.9) 44 (28.6) 127 (26.0)
757 (79.3) 379 (78.5) 1,136 (79.0) 226 (77.7) 24 (57.1) 110 (71.4) 360 (74.0)
.01
76 (28.0) 368 (21.5) 444 (22.3)
195 (72.0) 1,345 (78.5) 1,540 (77.7)
.02
.21
NOTE. African Americans were significantly more likely to have no third molar DFS compared with white subjects (P ⬍ .01). Subjects who were less well educated (basic [less than a high school education]) were significantly more likely to have no third molar DFS compared with those more highly educated (P ⫽ .04). Subjects who were former heavy or current tobacco users were significantly more likely to have no third molar DFS compared with those with lesser tobacco use (P ⫽ .01). Subjects with a diabetes diagnosis were significantly more likely to have no third molar DFS compared with those without a diabetes diagnosis (P ⫽ .02). Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
Discussion The most clinically relevant findings from these cross-sectional analyses of data from middle and olderaged subjects were as follows. Very few DARIC subjects, only 1% of the 2,003 DARIC subjects with visible third molars, were free of caries experience (ie, no DFS in any teeth including third molars). Seventyseven percent of subjects had coronal caries experience detected in at least 1 third molar. Caries experience was not unique to third molars: only 1% of subjects had DFS exclusively in third molars. Therefore almost all the subjects with third molar DFS also
had caries experience in more anterior teeth. Although fewer subjects had third molar DS overall compared with third molar FS, African Americans were significantly more likely to have third molar DS and less likely to have third molar FS compared with white subjects. Lower education level, lower income, or greater BMI were significantly associated with a greater number of subjects having third molar DS detected. Conversely, higher education level, higher income, and lower BMI were significantly associated with a greater number of subjects having third molar FS detected.
638
THIRD MOLAR CARIES
Table 3. MEAN (SD) DFS FOR DARIC SUBJECTS BY TOOTH TYPE COMPARED BY CORONAL CARIES EXPERIENCE IN AT LEAST 1 THIRD MOLAR (DFS) OR NONE (N ⴝ 2,003)
Mean DFS (SD) No Third Molar With DFS
ⱖ1 Third Molar With DFS
All remaining non–third molar teeth 12.1 (10.2) 19.4 (11.7) Second molars 2.6 (2.6) 4.8 (3.3) First molars 2.7 (3.1) 4.4 (3.8) Second premolars 2.6 (2.9) 4.1 (3.2) First premolars 2.2 (2.4) 3.3 (2.8) Canines 0.9 (1.6) 1.2 (1.8) Lateral incisors 0.9 (1.5) 1.2 (1.7) Central incisors 1.0 (1.6) 1.3 (1.8)
P Value ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001 ⬍.0001 .004 .0007 ⬍.0001
Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
How do the DARIC data on subjects retaining third molars compare with the few population data on the prevalence of visible third molars in the existing literature? Hugoson and Kugelberg9 have reported the only population data on the prevalence of third molars. The number of DARIC subjects with at least 1 visible third molar was lower than the number reported by Hugoson and Kugelberg for the cohort aged 60 years (31% vs 44%). However, in DARIC subjects with visible third molars we report a greater number of third molars compared with Hugoson and Kugelberg: 48% with only 1 third molar and 24% with at least 3 third molars compared with 21% and 12%, respectively. Not all third molars present in Hugoson and Kugelberg’s reported data were visible; some were covered, detected only by radiographs, and unlikely to have caries experience, further increasing the discrepancy in number of third molars and the prevalence of third molar caries experience in DARIC subjects. The 19% of subjects with visible third molars in the cohort aged 70 years reported by Hugoson and Kugelberg9 and the 31% of DARIC subjects were fewer than the 42% of 818 senior subjects in the analyses of Moss et al5 with visible third molars, who had a mean age of 73 years. However, we report a much higher prevalence of third molar caries experience in DARIC subjects than did Moss et al from senior subjects aged a decade older enrolled in the Piedmont 65⫹ population study (77% vs 28%). Caries experience was also more prevalent in teeth anterior to third molars in DARIC subjects with a visible third molar compared with the Piedmont 65⫹ subjects (98% vs 50%).5 In cross-sectional analyses of data from Piedmont 65⫹ and DARIC, third molar caries experience was significantly associated with caries experience in teeth more anterior in the mouth. This relationship alone may account for these differences.
Our data indicating that third molar caries experience was not exclusive to third molars is consistent with other reports. In the younger population of Shugars et al,6 with a mean age of 27 years, only 1% of 211 subjects had third molar caries experience without other teeth being affected. Moss et al5 reported that 8% of Piedmont 65⫹ subjects had caries experience exclusive to third molars; other retained teeth were not affected. Although these outcomes may be surprising to some clinicians, the findings are plausible biologically. Caries experience requires the colonization of specific, acidforming bacteria in biofilm attached to a non-sheddable surface of susceptible teeth. If the requisite micro-organisms are present in the biofilm on 1 tooth, all visible teeth in the mouth are at risk for a lifetime. Thus, for individuals with caries experience in teeth other than third molars earlier in life, the risk for third molar caries experience may be greater because third molars erupt over time. Conversely, if patients have no caries experience, third molar caries seems less likely. The prevalence of third molar caries experience may differ by ethnic group. A greater number of African-American DARIC subjects had at least 1 visible third molar than did white subjects (40% vs 29%), potentially increasing the risk for the prevalence of third molar caries experience in African Americans. However, we report that caries-free third molars (ie, no DS or FS) were significantly more prevalent in African Americans than white subjects. Moss et al10 reported similar results from the Piedmont 65⫹ subjects, an older population with a mean age of 73 years; white subjects were significantly more likely to have caries experience than African Americans (odds ratio, 2.5 [95% confidence interval, 1.4-4.3]). In our analyses, African Americans had a greater prevalence of third molar DS than white subjects. Drake et al11 reported similar findings from third molars and other remaining teeth in Piedmont 65⫹ subjects. Because half of the subjects studied in Piedmont 65⫹ were lifelong residents of a rural environment, Drake et al suggested that the differences were related to a lack of access to care. Our findings significantly associating lower income and lower education level with greater third molar DS could also be attributed to the same lack of access to care. The significant positive relationship between a history of diabetes and DS may be related to poorer overall health status, often complicated by less access to health care at some period in the affected individual’s life. The significant positive relationships we report between no use of tobacco or light former tobacco use or no diabetes diagnosis and having third molar FS was also reported by Moss et al10 from the Piedmont 65⫹ population; no tobacco use significantly increased the
639
FISHER ET AL
Table 4. DEMOGRAPHIC AND SOCIAL CHARACTERISTICS OF DARIC SUBJECTS WITH THIRD MOLAR CARIES EXPERIENCE (DS OR FS) IN AT LEAST 1 THIRD MOLAR OR NONE PER SUBJECT (N ⴝ 2,003)
No Third Molar ⱖ1 Third Molar No Third Molar ⱖ1 Third Molar With DS With DS P Value With FS With FS P Value Ethnicity/center [n (%)] African American Jackson, MS Forsyth County, NC Total White Forsyth County, NC Washington County, MD Suburbs of Minneapolis, MN Total Gender [n (%)] Male Female
277 (74.5) 40 (83.3) 317 (75.5)
95 (25.5) 8 (16.7) 103 (24.5)
206 (55.4) 21 (43.8) 227 (54.0)
166 (44.6) 27 (56.3) 193 (46.0)
470 (98.7) 448 (93.3) 595 (98.0) 1,513 (96.8)
6 (1.3) 32 (6.7) 12 (2.0) 50 (3.2)
⬍.01
107 (22.5) 96 (20.0) 112 (18.5) 315 (10.2)
369 (77.5) 384 (80.0) 495 (81.6) 1,248 (79.8)
⬍.01
1,039 (91.5) 809 (93.2)
96 (8.5) 59 (6.8)
.17
302 (26.6) 248 (28.6)
833 (73.4) 620 (71.4)
.33
Age [mean (SE)] (yr)
61.6 (5.5)
61.1 (5.1)
.35
61.3 (5.4)
61.6 (5.5)
.25
BMI [mean (SE)]
28.2 (4.9)
30.3 (6.3)
⬍.01
29.1 (5.2)
28.1 (5.0)
⬍.01
⬍.01
107 (45.3) 218 (5.4) 225 (24.9) 443 (25.0)
129 (54.7) 642 (74.7) 680 (75.1) 1,322 (75.0)
⬍.01
⬍.01
159 (40.5) 182 (26.0) 175 (21.3) 357 (23.4)
234 (59.5) 519 (74.0) 648 (78.7) 1,167 (76.6)
⬍.01
.08
247 (25.9) 123 (25.5) 370 (25.7) 71 (24.4) 23 (54.8) 52 (33.8) 146 (30.0)
708 (74.1) 360 (74.5) 1,068 (74.3) 220 (75.6) 19 (45.2) 102 (66.2) 341 (70.0)
.01
⬍.01
106 (39.1) 432 (25.2) 538 (27.1)
165 (60.9) 1,281 (74.8) 1,446 (72.9)
⬍.01
Education [n (%)] Basic (0-11 yr) Intermediate (12-16 yr) Advanced (ⱖ17 yr) Total Annual income [n (%)] Low (⬍$25,000) Medium ($25,000-$50,000) High (⬎$50,000) Total Tobacco use [n (%)] Never Former light Total Former heavy Current light Current heavy Total Diabetes diagnosis [n (%)] Yes No Total
189 (80.1) 801 (93.1) 856 (94.6) 1,657 (93.9)
47 (19.9) 59 (6.9) 49 (5.4) 108 (6.1)
319 (81.2) 660 (94.2) 792 (96.2) 1,452 (95.3)
74 (18.8) 41 (5.9) 31 (3.8) 72 (4.7)
880 (92.2) 445 (92.1) 1,325 (92.1) 279 (95.9) 36 (85.7) 140 (90.9) 455 (93.4)
75 (7.9) 38 (7.9) 113 (7.9) 12 (4.1) 6 (14.3) 14 (9.1) 32 (6.6)
229 (84.5) 1,604 (93.6) 1,833 (92.4)
42 (15.5) 109 (6.4) 151 (7.6)
Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
odds of detecting third molar caries experience (odds ratio, 2.3 [95% confidence interval, 1.1-4.4]). Although we cannot document a relationship between better overall health and FS, it is plausible that individuals who never or rarely use tobacco or do not have diabetes are in better health and may seek health care more frequently, including restorations for caries once it is detected. The association should be studied further in other populations. The findings we report from ARIC subjects have limits and may not be applicable to all middle-aged and older individuals. Although our study population numbers over 6,500 subjects, only 4 communities in the United States were represented. A greater percentage of African-American subjects was included
compared with the US population. Subjects who were not white or African American were excluded from analyses because few subjects were enrolled. Thus Asian and Latino subjects in the United States were not studied. Although few unerupted third molars would be affected by caries experience, the number of third molars in the ARIC subjects may be understated because radiographs to detect impacted teeth were not available to examiners. The few population studies documenting the presence or absence of third molars, including the data from ARIC, do not report whether missing third molars were removed to treat existing disease or to prevent future pathology, nor do the studies document the relationship between third molars and the number of other teeth that may
640
THIRD MOLAR CARIES
Table 5. ADJUSTED MEAN (SE) DS AND FS FOR DARIC SUBJECTS BY TOOTH TYPE COMPARED BY THIRD MOLAR CORONAL CARIES EXPERIENCE (DS OR FS) IN AT LEAST 1 THIRD MOLAR OR NONE (N ⴝ 2,003)
Adjusted Mean (SE)*
Adjusted Mean (SE)*
Tooth Type
No Third Molar With DS
ⱖ1 Third Molar With DS
P Value
No Third Molar With FS
ⱖ1 Third Molar With FS
P Value
Second molars First molars Second premolars First premolars Canines Lateral incisors Central incisors
0.08 (0.01) 0.09 (0.01) 0.09 (0.01) 0.05 (0.01) 0.03 (0.01) 0.04 (0.01) 0.04 (0.01)
0.63 (0.05) 0.65 (0.05) 0.44 (0.05) 0.25 (0.03) 0.22 (0.02) 0.31 (0.03) 0.28 (0.03)
⬍.01 ⬍.01 ⬍.01 ⬍.01 ⬍.01 ⬍.01 ⬍.01
2.69 (0.14) 2.84 (0.18) 2.65 (0.14) 2.28 (0.12) 0.94 (0.08) 0.97 (0.07) 0.98 (0.08)
4.72 (0.08) 4.40 (0.10) 4.09 (0.08) 3.25 (0.07) 1.16 (0.05) 1.16 (0.04) 1.21 (0.05)
⬍.01 ⬍.01 ⬍.01 ⬍.01 .02 .02 .02
*Adjusted for ethnicity/center, age, gender, tobacco use, education level, income level, BMI, and history of diabetes mellitus. Fisher et al. Third Molar Caries. J Oral Maxillofac Surg 2010.
be present or missing. As a consequence, our data on prevalence of caries experience in retained third molars and the association with caries experience overall may be understated or overstated. However, with these limitations, the data we report do add to our understanding of caries experience, a chronic infectious disease, and the relationship of third molar caries experience in retained third molars to caries experience in other more anterior teeth. As other populations are studied for caries experience, a clinical caries assessment of third molars should be included. Acknowledgments The ARIC study is being carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts N01-HC55015, N01-HC-55016, N01-HC-55018, N01-HC-55019, N01-HC-55020, N01-HC-55021, and N01-HC-55022. The authors thank the staff and participants of the ARIC study for their important contributions. This manuscript was approved by the ARIC Coordinating Committee.
References 1. Murdoch-Kinch CA, McLean ME: Minimally invasive dentistry. J Am Dent Assoc 134:87, 2003
2. Powell LV: Caries risk assessment: Relevance to the practitioner. J Am Dent Assoc 129:349, 1998 3. Powell LV: Caries prediction: A review of the literature. Community Dent Oral Epidemiol 26:361, 1998 4. Brown LJ, Wall TP, Lazar V: Trends in caries among adults 18 to 45 years old. J Am Dent Assoc 133:827, 2002 5. Moss KL, Beck JD, Mauriello SM, et al: Third molar periodontal disease and caries in senior adults. J Oral Maxillofac Surg 65:103, 2007 6. Shugars DA, Elter JR, Jacks TM, et al: Incidence of occlusal dental caries in asymptomatic third molars. J Oral Maxillofac Surg 63:341, 2005 7. Radike A: Criteria for the diagnosis of dental caries, in Proceedings of the Conference on Clinical Testing of Cariostatic Agents. Chicago, American Dental Association, 1968, p 88 8. Beck JD, Elter JR, Heiss G, et al: Relationship of periodontal disease to carotid artery intima-media wall thickness: The atherosclerosis risk in communities (ARIC) study. Arterioscler Thromb Vasc Biol 21:1816, 2001 9. Hugoson A, Kugelberg CF: The prevalence of third molars in a Swedish population. An epidemiological study. Community Dent Health 5:121, 1988 10. Moss KL, Beck JD, Mauriello SM, et al: Risk indicators for third molar caries and periodontal disease in senior adults. J Oral Maxillofac Surg 65:958, 2007 11. Drake CW, Beck JD, Lawrence HP, et al: Three-year coronal caries incidence and risk factors in North Carolina elderly. Caries Res 31:1, 1997