This study analyzed the relationship among general (sex, age, T-score, menopausal status), local factors (remaining teeth and periodontal status), and morphology of the mandibular cortex assessed by MCW or MCI. All factors together explained the variation in MCW and MCI in females more than in males.
T-score was the strongest predictor of MCW and MCI in females, while in males, T-score was the only significant factor associated with MCI. Previous studies mainly assessed the diagnostic efficacy of radiomorphometric indices for osteoporosis screening [9, 11]. Fewer studies explored the association between MCW, MCI, and BMD in females or males [13, 33, 34]. Analyzing the relationships between radiomorphometric indices and BMD and considering relevant confounding factors is necessary to support using these indices for osteoporosis screening, specifically in males, in whom such studies are lacking. Even though we cannot directly compare the odds ratios and β coefficients from other studies with ours due to different methods and regression models used, it is evident from previous research that females with lower skeletal BMD have thinner and more eroded cortexes [34–37]. A positive correlation between T-score and MCW in females ranging from 0.33 to 0.45 in previous studies was consistent with our findings [34–36]. Our study found no significant association between T-score adjusted for other factors and MCW in males. In contrast, two previous studies found an association of thin MCW in males with osteoporosis (T-score ≤ -2.5 SD) [13, 38]. Unlike our study, Leite et al. also found a correlation coefficient of 0.29 for MCW and BMD at the femoral neck in males [13]. MCI was associated with T-score in our study, and this finding was consistent with the study by Leite et al. [13], while another study did not find such an association [38]. This divergence in findings might be related to smaller study samples, differences in sampling procedures, or a lack of adjustment for other factors in previous studies.
Age was another major contributor to MCW and MCI in females but not males. Our results align with previous studies that found an interaction between age and sex in the way that cortical thickness reduced more prominently in females than males [15, 19, 20, 39]. Similar trends were observed in other studies exploring the bone geometry and cortical thickness at different skeletal sites in relation to age and sex [40, 41]. A plausible biological explanation for maintaining mandibular cortical thickness in older men could be sex hormones, which play a crucial role in bone formation and resorption. Estrogens increase endosteal and reduce periosteal bone formation in females during puberty, while androgens accelerate periosteal bone formation in growing males [4, 5]. These physiological mechanisms contribute to sex dimorphism in the adult skeleton. After a certain age, bone resorption exceeds bone formation at the inner bone surface in both sexes. However, bone formation continues at the outer bone surface faster in males than in females due to androgens. Thus, males maintain their cortical bone not because they lose less endosteal bone than women but due to a greater periosteal formation [5]. Our study also found that age was a significant predictor for MCI in females but not males (Table 2,3). On the contrary, several studies found that MCI became more eroded with age, regardless of gender [39, 42, 43]. At the same time, those studies did not consider other confounding factors, which may partly explain the disagreement between ours and previous findings. However, one of the studies was longitudinal, which strengthened their results compared to ours [43].
In our study number of remaining teeth was significantly associated with MCW and MCI in females but not males (Table 2,3). It is well established in previous research that mechanical strains and subsequent osteocyte response define the geometry and morphology of skeletal bones [44–46]. Thus, it would be appropriate to hypothesize that the lack of loading forces in edentulous people or those with fewer teeth can independently influence the mandibular cortex. Several studies have found an association between remaining teeth and mandibular cortical morphology, even when controlling for age [16, 19, 20, 42]. Okabe et al. found that the number of remaining teeth was weakly correlated with MCW equally for both sexes (0.19 male and 0.14 female) [14]. Dutra et al. found that the number of remaining teeth was related to the thickness of the mandibular cortex, irrespective of gender [20]. Unlike Okabe et al. and Dutra et al., we did not find relationships between remaining teeth and radiomorphometric indices in males. Our results were consistent with Taguchi et al., who also found no relationships between remaining teeth on both the upper and lower jaws and MCW in males [19]. However, it might be more logical to consider only mandibular teeth, like some previous studies, since we assess the mandibular cortex [16, 20]. Despite the significant association in females, remaining teeth contributed to the thin and eroded cortex to a minor extent. Similar to our study, Legerton et al. and Gulsahi et al. showed that the influence of dentition on cortical erosion and cortical thickness in the ante gonial region was weaker than that of age [16, 42].
In our study, menopausal status was significantly associated with MCW and MCI in univariate analysis but not after adjustments, meaning that the other factors confounded this association greatly. Unlike our results, two previous studies found an association between menopausal status and MCW and MCI but did not consider other relevant confounding factors [47, 48].
The association between periodontitis and mandibular cortical morphology is poorly studied. Recent systematic reviews have shown that many studies have explored the effect of osteoporosis on periodontal health [22, 49]. However, some researchers hypothesized the opposite relationship. Two longitudinal studies explored the independent effect of periodontal disease on skeletal bone tissue and found an increased risk of osteoporosis among people diagnosed with periodontitis after accounting for confounders [50, 51]. The rationale behind those hypotheses was that patients with periodontitis exhibit higher systemic levels of inflammatory mediators such as interleukin (IL-2, IL-6) and tumor necrosis factor (TNF-α) [24]. Those inflammatory mediators affect the remodeling of bone tissue and may thus increase the risk of osteoporosis development [52, 53]. One can also speculate that periodontitis may influence the mandibular cortex via systemic inflammation mechanisms. Tooth loss and mobility due to severe periodontitis leading to a lack of mechanical loading may also influence mandibular cortex morphology. Unlike few existing studies on the subject, this study found no relationship between mandibular cortical morphology (MCW and MCI) and periodontal status for females or males [47, 48]. That might be due to the difference in study design, the wide variation of periodontal measurements, and case definitions in the literature, which makes comparisons between studies difficult.
The current study has several limitations. First, it is a cross-sectional study. Despite the arguments regarding the possible links between sex, bone mineral density, age, remaining teeth, periodontitis, and mandibular cortex, we cannot infer the directions of these relationships. Another limitation is the suboptimal inter- and interobserver agreement of radiomorphometric indices found and reported in our previous work [32]. Moreover, the panoramic radiographs in this study were not standardized, i.e., we could not make precise corrections of MCW for magnification. Suboptimal observer agreement and image distortion due to the patient's head misalignment were likely to produce random errors and a substantial unexplained variation in mandibular cortical morphology.