Skeletal Class III malocclusion combines sagittal jaw disharmony with dentoalveolar compensation. Mandibular incisors in Class III patients often incline lingually, which can reduce the clinical severity of anterior crossbite but may position roots close to the cortical boundaries of the mandibular symphysis. Because incisor position influences decisions for camouflage treatment and presurgical decompensation, local root–alveolus relationships are clinically relevant.
Two-dimensional measures such as the incisor mandibular plane angle (IMPA) reference the mandibular plane rather than local alveolar housing, so similar IMPA values can coexist with different symphyseal widths, inclinations, and cortical distributions. When three-dimensional imaging is already clinically indicated, cone-beam computed tomography (CBCT) permits direct measurement of labial and lingual alveolar boundaries.
Previous CBCT studies have reported reduced mandibular anterior alveolar dimensions and increased dehiscence or fenestration in skeletal Class III subjects, with heterogeneity across sampled teeth, skeletal subgroups, and root levels. Recent work has emphasized that mean thickness values can mask the high proportion of sites where the cortical plate is minimally detectable. This study extends prior evidence by measuring all four mandibular incisors at three root levels in an untreated Class III cohort, reporting tooth-level frequency distributions and patient-level asymmetry, and by evaluating whether the signed angle between the incisor axis and the principal axis of the mandibular dentoalveolar bone (IA-PAMD) indicates apical root position within the symphysis.
The primary descriptive expectation was that labial bone would be thinner than lingual bone. A signed-angle analysis was added to distinguish root-apex eccentricity from differences in aggregate apical bone support.
This is a retrospective cross-sectional imaging study reported according to STROBE. Pretreatment wide-field CBCT and lateral cephalometric records acquired between June 2024 and June 2025 at the Department of Orthodontics, Hanoi National Hospital of Odonto-Stomatology (Hanoi, Vietnam) were screened. Records were accessed for research from June 1, 2025, to November 30, 2025. Only clinically indicated images were used; no additional imaging was performed for research. The study received Institutional Review Board approval (VNU University of Medicine and Pharmacy; study code 1820; May 16, 2025) and the requirement for individual informed consent was waived because records were de-identified prior to analysis.
The imaging archive initially contained 1,522 paired CBCT and cephalometric records. After re-screening with a revised final skeletal and clinical Class III definition (signed ANB < 0° plus clinical Class III canine or molar relationship and anterior edge-to-edge bite or crossbite), 80 patients met eligibility and were retained for analysis, yielding 320 mandibular incisors. Eligibility required pretreatment wide-field CBCT and lateral cephalogram acquired at the same diagnostic stage, 28 permanent teeth excluding third molars present on images, adequate image quality and a 13 x 13 cm field of view, complete apical closure of mandibular incisors, and the defined Class III diagnosis. Exclusion criteria included prior/ongoing orthodontic treatment, dental restorations or pathology affecting incisors, abnormal incisor morphology preventing standardized measurement, congenital craniofacial anomalies, previous orthognathic surgery, and maxillofacial trauma.
A post hoc study-size justification based on a previously reported correlation was reported during peer review; however, all eligible patients in the study period were included and no stopping rule influenced selection.
CBCT scans were obtained using a Planmeca ProMax 3D Max unit with a single acquisition protocol: 96 kV, 5.6 mA, 12.08-second scan time, 13 x 13 cm field of view, and 0.20-mm voxel size. DICOM data were imported into InVivo Dental 6/6.5. Images were oriented using the midsagittal plane and a horizontal plane based on bilateral orbital landmarks. Sagittal sections for each incisor were adjusted to pass through the crown and root center and confirmed in axial and coronal views prior to measurement.
Mandibular incisors were coded FDI 32, 31, 41, and 42. The long axis of each tooth was defined from the incisal midpoint to the root apex. Labial and lingual alveolar bone thicknesses were measured at three root levels: cervical, middle, and apical. The principal axis of the mandibular dentoalveolar bone (PAMD) and the angle between the incisor long axis and PAMD (IA-PAMD) were measured; IA-PAMD was reported as a signed angle to indicate whether the incisal portion of the tooth axis lies labial or lingual relative to the symphyseal axis. Statistical models adjusted for tooth, age, sex, signed ANB, GoGn-SN, and incisor mandibular plane angle.
Correlations and cluster-robust multivariable models were used to evaluate associations between signed IA-PAMD and apical bone distribution. Models reported effect estimates with 95% confidence intervals. Frequency distributions of cortical thickness below clinically relevant thresholds (0.5 and 1.0 mm) were described.
From 1,522 archived paired records screened during revision, 80 patients fulfilled the revised skeletal and clinical Class III criteria and were included. The analytic dataset comprised 320 mandibular incisors. The study reports that measurements were made on pretreatment CBCT images oriented and checked in three planes, and analyses adjusted for skeletal and dental covariates.
Mean labial thicknesses were reported as 0.36 mm at the cervical level, 0.49 mm at the middle level, and 2.40 mm at the apical level. Corresponding mean lingual thicknesses were 0.59 mm (cervical), 1.39 mm (middle), and 4.01 mm (apical). Across all comparisons, labial bone was thinner than lingual bone with statistical significance (P < 0.001).
Labial cortical thickness was below 0.5 mm in 80.3% of cervical observations and in 63.4% of middle-root observations, highlighting a high prevalence of minimally detectable labial cortex at coronal and mid-root levels in this untreated Class III cohort.
Signed IA-PAMD correlated positively with labial apical thickness and negatively with lingual apical thickness, but showed no correlation with total apical width. In cluster-robust multivariable models adjusted for tooth, age, sex, signed ANB, GoGn-SN, and incisor mandibular plane angle, each 1° increase in signed IA-PAMD was associated with a 0.084 mm greater labial apical thickness (95% CI, 0.053–0.115) and a 0.092 mm lower lingual apical thickness (95% CI, −0.141 to −0.043). The model estimate for total apical width was effectively null (−0.008 mm; 95% CI, −0.082 to 0.065).
The mirrored sign and similar magnitude of the labial and lingual coefficients, along with the near-zero total-width coefficient, support interpreting signed IA-PAMD as indicating root-apex eccentricity within the symphysis rather than representing greater aggregate apical bone support.
This CBCT study of untreated skeletal Class III patients confirms that mean labial alveolar bone thickness around mandibular incisors is thinner than lingual thickness and that a large proportion of coronal and mid-root labial measurements are below 0.5 mm. By assessing all four mandibular incisors at three root levels and reporting both frequency distributions and adjusted associations, the study adds granularity to prior reports that emphasized reduced anterior alveolar dimensions in Class III malocclusion.
The analysis of signed IA-PAMD indicates that variation in that angle reflects displacement of the root apex within the symphysis (eccentricity) rather than overall apical bone volume, because increases in signed IA-PAMD were associated with reciprocal changes in labial and lingual apical thickness but not with total apical width. Therefore, signed IA-PAMD may serve as a descriptor of root position relative to the local alveolar axis, complementing global inclination measures such as IMPA which reference the mandibular plane.
The authors emphasize that cross-sectional associations should not be interpreted as biological limits for orthodontic movement. Specifically, signed IA-PAMD should not be treated as a therapeutic target or a safety threshold for tooth movement without further evidence. Additionally, the study lacked a Class I comparator group, which limits claims that the findings are specific to skeletal Class III morphology.
In this retrospective cross-sectional CBCT study of 80 untreated skeletal Class III patients (320 mandibular incisors), labial alveolar bone was significantly thinner than lingual bone at cervical, middle, and apical levels, with labial thickness often below 0.5 mm at coronal and mid-root measurements. Signed IA-PAMD was associated with reciprocal changes in labial and lingual apical thickness but not with total apical width, supporting its interpretation as a measure of root-apex eccentricity within the symphysis rather than aggregate apical support. The authors caution against using signed IA-PAMD as a treatment target or safety threshold and note that the absence of a Class I reference group limits morphology-specific conclusions.