Age and Third Molar Extraction as Risk Factors for Temporomandibular Disorder

Abstract

This study investigated third molar removal as a risk factor for temporomandibular disorder (TMD) in all age groups. We compared 2217 Kaiser Permanente Northwest health plan enrollees with a history of third molar extraction with 2217 age-and gender-matched enrollees with radiographic confirmation of no lifetime third molar removal. Common Dental Terminology codes were used to identify information on third molar removal, and International Classification of Disease codes were used to identify TMD. Relative risks were calculated overall, and by each decade of life, in univariate and multivariate analyses. The incidence of TMD in subjects with and without third molar removal were 7 and 5 per thousand person-years, respectively. Third molar removal among subjects of all ages resulted in a statistically insignificant increased relative risk for TMD (1.4, 95% confidence interval (CI): 0.9–2.2). The relative risk was slightly higher in those under 21, but was also not statistically significant (1.6, CI: 0.8–3.1).

INTRODUCTION

In a recent retrospective cohort study, third molar removal was found to be associated with a significantly increased incidence of temporomandibular disorder (TMD) (Huang and Rue, 2006). That study utilized dental insurance claims data for ascertainment of third molar removal and subsequent TMD status, and was restricted to a population of 15- to 20-year-olds. This is a period when many third molars are unerupted or partially erupted. It is not uncommon, however, for third molars also to be extracted in older individuals, after these teeth have had a continued opportunity to erupt. No literature is available on the relationship between age at time of third molar extraction and subsequent risk for TMD. Thus, this study investigated third molar removal as a risk factor for TMD among all age groups.

MATERIALS & METHODS

The participants in this population-based cohort study were drawn from enrollees of Kaiser Permanente Northwest (KPNW). KPNW is a group-model Health Maintenance Organization that provides complete medical and dental services to a population of about 200,000 individuals in Southwest Washington and Northwest Oregon. Since its inception, KPNW has used a unique health record number, enabling researchers to link data on all health care that an individual has received under the Kaiser system. Third molar extraction is covered under the dental program, and TMD care is covered under both medical and dental programs. This study was approved by the Institutional Review Board at the University of Washington, as well as by the KPNW Institutional Review Board.

Exposed Cohort

Initially, all individuals were identified who had continuous enrollment in KPNW during the study period of January 1, 1998, to December 31, 2003. From this population, all those who had one or more third molars removed during the study period were considered for inclusion in the exposed cohort. These persons were identified by tooth numbers 1, 16, 17, and 32, in conjunction with Current Dental Terminology (CDT) extraction codes 7110–7241. Any of these individuals who had TMD diagnoses in the first 3 mos of the study period or who did not have at least 1 yr of follow-up after removal of third molars were excluded.

Unexposed Cohort

From the same initial population, three gender- and age-matched individuals (within 1 yr) who did not have third molars removed during the study period were identified for each person who had third molar removal. Because these comparison individuals might have undergone third molar removal prior to the start date of this study, dental radiographs for these persons were reviewed to verify the presence of all 4 third molars during the study period, or that third molars were congenitally missing. For each exposed person, we sequentially reviewed the three potentially unexposed persons to determine if one met these criteria. The outcome of interest in this study was the first indication of temporomandibular disorder (TMD), as identified according to the International Classification of Disease, 9th edition (ICD-9), diagnostic codes 524.6 through 524.69.

Statistical Analyses

The data were analyzed in a cohort design, with one-to-one matching of exposed and unexposed persons. Matched pairs in which TMD occurred for either individual prior to the date of third molar removal (the reference date) were excluded from the analyses. In each matched pair, ‘person-time’ accrued from the reference date until the end of the study, or until the time that an individual received a TMD diagnosis. In those situations, person-time was no longer accrued for the person who received a TMD diagnosis. The relative risk was calculated with incidence densities for the exposed and unexposed cohorts. We used Poisson regression to obtain relative risk estimates associated with extraction adjusted for dental utilization. Dental utilization was based on the annual rate of dental prophylaxis visits, based on the CDT codes 1110 and 1120. Among exposed individuals, Poisson regression was also used to explore the multivariate associations between TMD and extraction characteristics, including provider, severity of impaction, type of anesthesia, and arch of extraction. ‘Provider’ was coded as either oral surgeon or general dentist. ‘Severity of impaction’ was classified based on the following codes:

Level 1: 7110, 7120, and 7130—routine extraction of erupted tooth

Level 2: 7210 and 7220—extraction of erupted or impacted tooth requiring mucoperiosteal flap

Level 3: 7230—extraction of impacted tooth, partial bony

Level 4: 7240—extraction of impacted tooth, completely bony

Level 5: 7241 and 7250—extraction of impacted tooth, completely bony, with unusual surgical complications

Anesthesia was categorized as local anesthesia only, local anesthesia with nitrous oxide, intravenous sedation, and general anesthesia, based on CDT codes. Arch of extraction was coded as maxillary only, mandibular only, or both. All statistical analyses were performed with SAS version 9.1 (Cary, NC, USA). A significance level of P < 0.05 was established for all tests.

RESULTS

The selection of the sample is outlined in the Fig. We confirmed 2296 matched pairs, 79 of which were excluded due to TMD outcomes prior to the date of extraction, leaving a total of 2217 pairs for analysis. The distribution of TMD diagnoses is provided in Table 1 . The largest group consisted of unspecified temporomandibular joint (TMJ) disorders, followed by arthralgia, articular disc disorders, and other specified TMJ disorders. While the exposed persons had proportionately more arthralgia, articular disc disorders, and specified disorders, these differences were not statistically significant.

Of the 2217 matched pairs, about 50% of the participants fell in the 10- to 20-year-old age bracket, 25% in the 21 to 30 age bracket, and progressively fewer in each subsequent decade. About 64% of all participants were male, and 36% were female. However, of the 77 participants with TMD diagnoses, 50 were female. Thus, female gender was associated with a significantly increased relative risk of TMD (Table 2 ). The annual incidence of TMD in the exposed and unexposed cohorts was 0.7% and 0.5%, respectively. The annual incidence was much higher in females (1.1%) compared with males (0.3%). Dental utilization was 0.3 prophylaxis visits/year in both cohorts.

For the entire sample, the relative risk for TMD in individuals who had third molars removed compared with that in individuals who did not was 1.4 (Table 2 ). Although the point estimate was elevated, this increase was not statistically significant. When stratified by age, the point estimates for the relative risks were 1.6 in the 15- to 20-year-olds, 1.1 in the 21-to 30-year-olds, 1.2 in the 31- to 40-year-olds, and 2.5 in the participants over 40. However, none of these relative risks reached statistical significance. The estimates were minimally affected by adjustment for utilization.

Among participants with third molar extraction, clear patterns emerged for parameters related to the removal of teeth with regard to age (Table 3 ). With each increasing decade of life, participants were less likely to have all 4 third molars extracted at one visit, and the method of anesthesia utilized became progressively simpler. Additionally, third molars became increasingly accessible, as indicated by the CDT codes, which described the relative degree of eruption or impaction. Finally, while the majority of third molar removal was performed by oral surgeons through the age of 20, general dentists performed more than half of the third molar extractions after the age of 20, and this percentage increased to over 75% by age 50. Although there were no significant relationships between the extraction parameters and subsequent risk for TMD in univariate or multivariate analyses, analysis of the data suggested a relationship between severity of impaction and TMD. An analysis of the yearly time intervals following third molar removal indicated that the risk for TMD was elevated during the first year after third molar removal (Table 2 ).

DISCUSSION

Third molar removal has been implicated as a risk factor for TMD in several descriptive TMD papers (Greene et al., 1969; Butler et al., 1975; Pullinger et al., 1988; Huang et al., 2002). Additionally, three studies specifically investigated the relationship between third molar removal and TMD. Raustia compared 22 university students before and after third molar removal, and found an increased severity of TMJ dysfunction 3 mos after third molar removal (Raustia and Oikarinen, 1991). Threlfall conducted a case-control study, utilizing 220 TMD cases and 1100 age- and gender-matched control individuals. He reported an odds ratio of 1.28 (95% CI, 0.96–1.71) in a mostly older population (Threlfall et al., 2005). Finally, Huang reported a significantly elevated risk of 1.6 for TMD after third molar removal in a sample of 35,000 15- to 20-year-olds (Huang and Rue, 2006).

The current study did not find significantly elevated risks for TMD subsequent to third molar extraction. However, the point estimate for individuals under the age of 21 does suggest a relationship, especially in light of the identical risk estimate from Huang’s prior work in a completely independent sample (Huang and Rue, 2006). In both studies, the relative risk of 1.6 for teens and young adults after third molar removal was only moderately increased, and the overall incidence of TMD under the age of 21 appeared to be relatively low (around 1%). Although TMD is not common in this age group, the high frequency of third molar removal results in a population-attributable risk of 23%, indicating that almost a quarter of all TMD cases in this age group might be related to third molar extraction.

The patterns related to third molar removal in younger persons may partially explain the elevated risk, since they were more likely to have all 4 third molars removed at one time, deeper levels of anesthesia, and more severely impacted third molars. Oral surgeons often advocate third molar removal during the teens, citing advantages such as incompletely formed roots, better healing, and less morbidity (Bruce et al., 1980; Osborn et al., 1985; de Boer et al., 1995; Phillips et al., 2003). However, these studies typically obtain participants from oral surgery practices, and difficult extractions among older individuals may be over-represented in these populations. Studies that do not account for this factor are likely to overestimate the impact of age (Phillips et al., 2003). In fact, others, using prospective designs or multivariate analyses, have reported that the risk of complications is not increased greatly in older age groups (Bui et al., 2003; Haug et al., 2005). The non-significant relative risks of 1.1 and 1.2 for persons in their third and fourth decades of life are similar to the odds ratio of 1.3 reported in a prior study (Threlfall et al., 2005). The consistency of these estimates would suggest that third molar removal is not likely to be a large risk factor for TMD in older populations. This might be explained by the simpler extraction patterns observed in older persons.

While this study did not find a significant relationship between severity of impaction and risk for TMD, a tendency for this to occur was observed. Severity of impaction has also been described as being associated with increased risk for other complications. For instance, alveolar osteitis, secondary infection, and dysesthesia were all more common when impacted teeth were removed (10%, 10%, and 1.2%, respectively), compared with erupted teeth (2%, 0.2%, and 0.2%, respectively) (Osborn et al., 1985). It has been reported that third molars can continue to erupt in the 3rd and 4th decades of life, and this eruption is difficult to predict (Hattab, 1997; Kruger et al., 2001; Venta et al., 2004; Nance et al., 2006). On the basis of this information, simply delaying removal of asymptomatic third molars until the mid- to late 20s might reduce extraction-related morbidity.

Obviously, the decision to extract third molars is complex, and can involve many factors, such as potential health benefits, risks and complications, family preferences, professional recommendations, and insurance coverage. While three cost-effectiveness analyses have concluded that retention of asymptomatic third molars is the most cost-effective approach with the least disability (Tulloch et al., 1990; Edwards et al., 1999; Song et al., 2000), third molars continue to be the most commonly extracted teeth in teenagers and young adults.

A clear inverse relationship was seen between age and severity of impaction, with 74% of the extractions done prior to the age of 21 reported as full bony impactions, compared with about 5% after the age of 40. These age-related extraction patterns could be influenced by several factors. Continued eruption of third molars is one explanation. Additionally, more severe impactions might be preferentially referred earlier in life, resulting in an increased prevalence of third molar extractions with bony impactions in teenagers. Also, it is possible that care-seeking behavior differs over time. The current study raises interesting questions about age and risk for complications, since it included persons who had extractions from all dental providers. The patterns we found for severity of impaction by age were quite different from those reported in the studies that enrolled individuals from only oral surgery practices, and these relationships merit further investigation.

A strength of this study was its ability to verify that comparison individuals had no lifetime history of third molar removal, thus preventing misclassification of the exposure. The central chart-housing facility in which all Kaiser dental charts are maintained made this verification process feasible. Only about 20% of the potential comparison individuals could be confirmed as having no lifetime third molar removal. Although this figure may seem low, from 40 to 50% of individuals in insured populations have third molars removed by the age of 20 (Eklund and Pittman, 2001; Huang and Rue, 2006), and the rate of third molar agenesis has been reported to be between 9 and 35% (Thompson et al., 1974). Other strengths of this study were its population-based cohort design, the age- and gender-matched sample, largely equal access to dental and TMD care by both cohorts, and the unbiased ascertainment of exposure and outcome information. Additionally, individuals who had extractions from all dental providers were included. One limitation was the relatively low incidence of TMD, which affected the power of the statistical analyses. Our initial estimates for TMD rates were 3% and 4.5% in the unexposed and exposed groups, respectively, which would have resulted in 80% power. Post hoc calculations indicated that the power of our study to detect a statistically significant difference was only 34%. The results of this study should be generalizable to other similarly insured populations in the US.

Based on the results of this study, extraction patterns exhibit considerable change over the decades of life. The risk for TMD after third molar removal was not statistically elevated overall, or for each decade of life. However, analysis of these data suggests that the risk may be greater for persons under the age of 21, at which time the factors associated with third molar removal indicate that it is usually a more invasive procedure.

Table 1.

Distribution of TMJ Diagnoses

		Unexposed	Exposed
524.6	TM joint disorders, unspecified	17	14
524.62	Arthralgia of TM joint	8	17
524.63	Articular disc disorder	5	8
524.69	Other specified TM joint disorder	2	6

Table 2.

Relative Risks for TMD

		TMD Events	Person-years	TMD Incidence	Relative Risk (95% Confidence limits)
* Significant at the 0.05 level.
Gender
Female		50	4698	0.011	3.5 (2.2, 5.6)
Male		27	8850	0.003
Third molars extracted
Yes (exposed)		45	6753	0.007	1.4 (0.9, 2.2)
No (unexposed)		32	6795	0.005
Reference age
9–14 yrs	Exposed	0	53		—
	Unexposed	0	53
15–20 yrs	Exposed	22	2874	0.008	1.6 (0.8, 3.1)
	Unexposed	14	2889	0.005
21–30 yrs	Exposed	10	1710	0.006	1.1 (0.5, 2.7)
	Unexposed	9	1718	0.005
31–40 yrs	Exposed	8	1021	0.008	1.2 (0.4, 3.2)
	Unexposed	7	1037	0.007
41 50 yrs	Exposed	3	619	0.005	3.0 (0.3, 29)
	Unexposed	1	619	0.002
51 60 yrs	Exposed	2	372	0.005	2.0 (0.2, 22)
	Unexposed	1	376	0.003
61 94 yrs	Exposed	0	103	0.000	—
	Unexposed	0	103	0.000
Yrs following extraction
1st	Exposed	20	2208	0.009	2.2* (1.0, 4.9)
	Unexposed	9	2213	0.004
2nd	Exposed	9	1883	0.005	0.9 (0.4, 2.2)
	Unexposed	10	1895	0.005
3rd	Exposed	10	1342	0.007	1.7 (0.6, 4.6)
	Unexposed	6	1355	0.004
4th	Exposed	3	835	0.004	0.6 (0.1, 2.5)
	Unexposed	5	846	0.006
Oral surgeon
Yes		22	3103	0.007	1.1 (0.6, 2.0)
No		23	3650	0.006
Type of anesthesia
Local anesthesia		21	3568	0.006	1.0 (0.7, 1.5)
Nitrous oxide/conscious sedation		21	2566	0.008
Intravenous sedation		3	454	0.007
General anesthesia		0	164	0.000
Location of extractions
Maxillary arch only		10	1390	0.007	1.0 (0.7, 1.4)
Mandibular arch only		8	1335	0.006
Both arches		27	4028	0.007
Difficulty of extractions
Erupted tooth, routine		13	1966	0.007	1.2 (0.9, 1.5)
Erupted tooth or soft-tissue impaction, surgical		5	1489	0.003
Partial bony impaction, surgical		16	2133	0.008
Complete bony imapction, surgical		10	1108	0.009
Complete bony impaction, surgical w/complications		1	57	0.018

Table 3.

Extraction Parameters by Age

	Age at Extraction
	9–20		21–30		31–40		41–50		51–60		61–94		Totals
^a Subgroups with the largest percentage of participants are in bold.
Provider	n	%	n	%	n	%	n	%	n	%	n	%	n	%
General dentist	393	36	280	56 ^a	205	71	149	78	89	75	25	83	1141	51
Oral surgeon	696	64	220	44	85	29	41	22	29	25	5	17	1076	49
Anesthesia
Local	332	30	288	58	214	74	165	87	104	88	28	93	1131	51
Nitrous	585	54	177	35	65	22	19	10	11	9	2	7	859	39
IV sedation	114	10	24	5	7	2	4	2	3	3	0	0	152	7
General anesthesia	58	5	11	2	4	1	2	1	0	0	0	0	75	3
Arch of extraction
Maxilla	43	4	114	23	128	44	81	43	43	36	11	37	420	19
Mandible	144	13	89	18	68	23	65	34	54	46	17	57	437	20
Both	902	83	297	59	94	32	44	23	21	18	2	7	1360	61
Severity of impaction
Erupted	73	7	148	30	145	50	125	66	68	58	21	70	580	26
Soft tissue	168	15	119	24	87	30	49	26	38	32	8	27	469	21
Partial bony	568	52	155	31	36	12	12	6	5	4	1	3	777	35
Complete bony	276	25	75	15	18	6	3	2	4	3	0	0	376	17
Complete bony w/complications	4	0	3	1	4	1	1	1	3	3	0	0	15	1
Total participants (n)	1089		500		290		190		118		30		2217

Figure.

Selection of sample.

Footnotes

Notes

Acknowledgements

This study was supported by NIDCR grant #DE14609-01. The authors thank Drs. Linda LeResche, Alex White, Gerardo Maupome, Kari Borgen, and David Covell. We also thank Mr. Paul Cheek and the Kaiser Permanente Center for Health Research for their assistance with this project.

References

Bruce RA, Frederickson GC, Small GS -1980-. Age of patients and morbidity associated with mandibular third molar surgery. J Am Dent Assoc 101:240–245.

Bui CH, Seldin EB, Dodson TB -2003-. Types, frequencies, and risk factors for complications after third molar extraction. J Oral Maxillofac Surg 61:1379–1389.

Butler JH, Folke LE, Bandt CL -1975-. A descriptive survey of signs and symptoms associated with the myofascial pain-dysfunction syndrome. J Am Dent Assoc 90:635–639.

de Boer MP, Raghoebar GM, Stegenga B, Schoen PJ, Boering G -1995-. Complications after mandibular third molar extraction. Quintessence Int 26:779–784.

Edwards MJ, Brickley MR, Goodey RD, Shepherd JP -1999-. The cost, effectiveness and cost effectiveness of removal and retention of asymptomatic, disease free third molars. Br Dent J 187:380–384.

Eklund SA, Pittman JL -2001-. Third-molar removal patterns in an insured population. J Am Dent Assoc 132:469–475.

Greene CS, Lerman MD, Sutcher HD, Laskin DM -1969-. The TMJ pain-dysfunction syndrome: heterogeneity of the patient population. J Am Dent Assoc 79:1168–1172.

Hattab FN -1997-. Positional changes and eruption of impacted mandibular third molars in young adults. A radiographic 4-year follow-up study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 84:604–608.

Haug RH, Perrott DH, Gonzalez ML, Talwar RM -2005-. The American Association of Oral and Maxillofacial Surgeons Age-Related Third Molar Study. J Oral Maxillofac Surg 63:1106–1114.

10.

Huang GJ, Rue TC -2006-. Third-molar extraction as a risk factor for temporomandibular disorder. J Am Dent Assoc 137:1547–1554.

11.

Huang GJ, LeResche L, Critchlow CW, Martin MD, Drangsholt MT -2002-. Risk factors for diagnostic subgroups of painful temporomandibular disorders -TMD-. J Dent Res 81:284–288.

12.

Kruger E, Thomson WM, Konthasinghe P -2001-. Third molar outcomes from age 18 to 26: findings from a population-based New Zealand longitudinal study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 92:150–155.

13.

Nance PE, White RP Jr, Offenbacher S, Phillips C, Blakey GH, Haug RH -2006-. Change in third molar angulation and position in young adults and follow-up periodontal pathology. J Oral Maxillofac Surg 64:424–428.

14.

Osborn TP, Frederickson G Jr, Small IA, Torgerson TS -1985-. A prospective study of complications related to mandibular third molar surgery. J Oral Maxillofac Surg 43:767–769.

15.

Phillips C, White RP Jr, Shugars DA, Zhou X -2003-. Risk factors associated with prolonged recovery and delayed healing after third molar surgery. J Oral Maxillofac Surg 61:1436–1448.

16.

Pullinger AG, Seligman DA, Solberg WK -1988-. Temporomandibular disorders. Part II: Occlusal factors associated with temporomandibular joint tenderness and dysfunction. J Prosthet Dent 59:363–367.

17.

Raustia AM, Oikarinen KS -1991-. Effect of surgical removal of the mandibular third molars on signs and symptoms of temporomandibular dysfunction: a pilot study. Cranio 9:356–360.

18.

Song F, O’Meara S, Wilson P, Golder S, Kleijnen J -2000-. The effectiveness and cost-effectiveness of prophylactic removal of wisdom teeth. Health Technol Assess 4-15-:1–55.

19.

Thompson GW, Popovich F, Anderson DL -1974-. Third molar agenesis in the Burlington Growth Centre in Toronto. Community Dent Oral Epidemiol 2:187–192.

20.

Threlfall AG, Kanaa MD, Davies SJ, Tickle M -2005-. Possible link between extraction of wisdom teeth and temporomandibular disc displacement with reduction: matched case control study. Br J Oral Maxillofac Surg 43:13–16.

21.

Tulloch JF, Antczak-Bouckoms AA, Ung N -1990-. Evaluation of the costs and relative effectiveness of alternative strategies for the removal of mandibular third molars. Int J Technol Assess Health Care 6:505–515.

22.

Venta I, Ylipaavalniemi P, Turtola L -2004-. Clinical outcome of third molars in adults followed during 18 years. J Oral Maxillofac Surg 62:182–185.