Abstract
The prophylactic removal of asymptomatic third molars is a common but controversial procedure often rationalized as necessary to prevent future disease on adjacent teeth. Our objective in this retrospective cohort study of adult men was to examine whether second-molar loss differed by baseline status of the adjacent third molar, taking into account the individual's overall state of oral hygiene, caries, and periodontitis. We analyzed data from participants of the VA Dental Longitudinal Study who had at least 1 second molar present at baseline and 2 or more triennial dental examinations between 1969 and 2007. We classified second molars by third-molar status in the same quadrant: unerupted, erupted, or absent. Tooth loss and alveolar bone loss were confirmed radiographically. Caries and restorations, calculus, and probing depth were assessed on each tooth. We estimated the hazards of second-molar loss with proportional hazards regression models for correlated data, controlling for age, smoking, education, absence of the first molar, and whole-mouth indices of calculus, caries, and periodontitis. The analysis included 966 men and 3024 second molar/first molar pairs. Follow-up was 22 +/- 11 y (median 24, range 3-38 y). At baseline, 163 third molars were unerupted, 990 were erupted, and 1871 were absent. The prevalence of periodontitis on the second molars did not differ by third-molar status. The prevalence of distal caries was highest on the second molars adjacent to the erupted third molars and lowest on the second molars adjacent to the unerupted third molars. Relative to the absent third molars, adjusted hazards of loss of second molars were not significantly increased for those adjacent to erupted (hazard ratio [HR] = 0.96, 95% confidence interval [CI] = 0.79-1.16) or unerupted (HR = 1.25, 95% CI = 0.91-1.73) third molars. We found similar results when using alveolar bone loss as the periodontitis indicator. Our findings suggest that retained third molars are not associated with an increased risk of second-molar loss in adult men.