LAUSR

Background Prevalence of distal caries in mandibular second molars (M2Ms) and its relationship with impacted condition of the adjacent mandibular third molars (M3Ms) have been reported in some studies. The results, however, were ambiguous because of including all impaction types and using univariate analysis for statistics. Aim This study aimed to determine anatomical features of mesially/horizontally impacted mandibular third molars (M3Ms) that could predict distal caries in the adjacent mandibular second molars (M2Ms) using multivariable analysis. Materials and Methods The study sample consisted of 300 digital panoramic radiographs of patients who underwent impacted M3Ms extraction. Two independent researchers collected the following variables from 446 pairs of M2M-M3M: sex, age, status of distal caries in M2Ms, mesial angulation, and Pell–Gregory classification of M3Ms. Results The prevalence of distal caries was 50.67%. Multivariable Firth's logistic regression analysis showed that age (β = 0.066, 95% CI = 0.023–0.113), mesial angulation (<30°: β = −1.205, 95% CI = −1.955 to −0.499; >70°: β = −0.730, 95% CI = −1.184 to −0.282), vertical position (level B: β = 2.275; 95% CI = 0.015–7.175; level A: β = 3.008; 95% CI = 0.755–7.905), and horizontal position (level II: β = 1.515; 95% CI = 0.444–2.874; level I: β = 1.423; 95% CI = 0.283–2.825) were significant variables after adjusting for sex in the final model for predicting distal caries (p < 0.05). Conclusions In conclusion, anatomical positions of impacted M3Ms, such as mesial angulation and Pell–Gregory classification were significant predictors of distal caries in M2Ms.