“…According to the validation test using the categorical periodontal classification, we suggest that this classification tends to have a wide range of periodontal status in the moderate group so that some of the subjects who answered negative for our questionnaire could have periodontal condition which is very close to mild, but classified moderate. However, our study as well as a number of previous studies has demonstrated that combinations of the two questions used in this study, namely loose teeth and professional diagnosis of gum disease, can achieve high positive predictive value for moderate to severe periodontitis (Eke et al., ; Page & Eke, ), although rather lower sensitivity rates suggest that while the self‐reported periodontitis measure used here is valid, it is likely to markedly underestimate the total number of actual cases of periodontitis in this cohort (Chatzopoulos, Tsalikis, Konstantinidis, & Kotsakis, ; Dietrich et al., ; Eke et al., ; Heaton et al., ) and compound reporting biases. However, there is little information of the accuracy of self‐reported bleeding, and thus, the phenotype of “gum bleeding” may be less robust than the presence of self‐reported periodontitis.…”