Retrospective CBCT-based anatomical assessment of mandibular first and second molars: a cross-sectional study with potential implications for immediate implant planning

Abstract

Immediate implant placement in mandibular molar sockets can shorten treatment and reduce the number of surgical interventions, but the anatomy of the posterior mandible may limit implant positioning and increase surgical risk. This study evaluated six cone-beam computed tomography (CBCT)-derived anatomical parameters in mandibular first and second molars. This retrospective cross-sectional study included 25 CBCT scans from 25 adults, providing 100 molar sites (#36, #37, #46, and #47). Buccal bone thickness, crestal bone width, crestal bone height loss, vertical bone height, root apex-to-inferior alveolar canal (IAC) distance, and interradicular distance were measured. The data were analyzed using 2 × 2 repeated-measures analyses of variance with molar type and side as within-subject factors. Pairwise comparisons were Holm-adjusted. Inter-examiner reliability was evaluated using intraclass correlation coefficients. Inter-examiner reliability was excellent for all parameters (ICC range, 0.907–0.988). Second molars had greater buccal bone thickness (mean difference, 0.352 mm; 95% CI, 0.186–0.518; p < 0.001) and crestal bone width (0.786 mm; 95% CI, 0.367–1.205; p = 0.001). First molars had greater vertical bone height (1.724 mm; 95% CI, 1.050–2.398; p < 0.001), root apex-to-IAC distance (0.802 mm; 95% CI, 0.261–1.343; p = 0.005), and interradicular distance (0.924 mm; 95% CI, 0.655–1.193; p < 0.001). Crestal bone height loss did not differ significantly by molar type. Mandibular first and second molars demonstrated distinct CBCT-derived anatomical patterns in this selected dentate sample. These measurements may be informative for preoperative anatomical assessment. Their value for selecting implant dimensions or osteotomy positions, or predicting clinical outcomes, was not evaluated.

Clinical trial registration

Not applicable; this study did not involve a prospective clinical trial or patient intervention.

Introduction

Immediate implant placement (IIP) in molar extraction sockets is an established treatment approach that can reduce overall treatment duration, limit the number of surgical interventions, and help preserve post-extraction alveolar ridge architecture. A systematic review and meta-analysis by Atieh et al. found that immediate molar implant placement may achieve outcomes comparable to conventional delayed implant placement performed after socket healing, provided that appropriate case selection and sound surgical principles are followed (Atieh et al. 2010). Ragucci et al. similarly reported high survival rates for implants placed immediately in molar extraction sockets, although marginal bone remodeling and site-specific clinical factors must be considered (Ragucci et al. 2020). More recently, Witoonkitvanich et al. demonstrated high primary stability for computer-guided immediate implants placed in fresh maxillary and mandibular molar sockets (Witoonkitvanich et al. 2025).

Despite these favorable findings, IIP in the posterior mandible remains technically demanding because of the complex relationship among the extraction socket, interradicular septum, cortical plates, and inferior alveolar canal (IAC) (Ebrahimi Dastgurdi et al. 2025, Ho et al. 2022, Chen et al. 2021, Behfarnia et al. 2025). These anatomical characteristics may restrict implant dimensions and angulation, compromise primary stability, or increase the risk of lingual cortical perforation, IAC encroachment, and potential inferior alveolar nerve (IAN) injury. ^4,7,19,26 Consequently, CBCT has become an essential component of preoperative assessment because it permits three-dimensional evaluation of site-specific anatomy and facilitates prosthetically driven implant planning. ^4,5,18,20 Geng et al. demonstrated that CBCT-based dynamic navigation increased the safety distance from the IAC and reduced angular and positional deviations during immediate implant placement in the posterior mandible (Geng et al. 2024). A randomized clinical trial published in the Saudi Dental Journal also showed that computer-guided surgery improved implant-placement accuracy compared with conventional freehand placement, particularly in anatomically restricted ridges (Mahmoud et al. 2024).

The thickness and morphology of the surrounding alveolar bone are important considerations when assessing a molar extraction socket. ^6,7,8 A systematic review by Heimes et al. demonstrated substantial variation in buccal bone thickness according to tooth type and site, indicating that posterior cortical thickness should not be assumed without individual radiographic assessment (Heimes et al. 2021). In a recent CBCT analysis of more than 400 mandibular molar sites, Ebrahimi Dastgurdi et al. found greater risks of lingual cortical perforation and IAC encroachment at second-molar sites, whereas first molars more frequently provided adequate interradicular bone for immediate implant placement (Ebrahimi Dastgurdi et al. 2025). Shi et al. likewise reported that mandibular first molars generally provided adequate interradicular bone height, but the available interradicular width was often limited and the buccal plate was relatively thin, particularly adjacent to the mesial root (Shi et al. 2024).

The relationship between mandibular molar roots and the IAC is another major determinant of surgical safety (Bürklein et al. 2015). Bürklein et al. reported that second-molar root apices were positioned significantly closer to the mandibular canal than those of first molars in a German population (Bürklein et al. 2015). Similar tooth-specific variation was documented by Aljarbou et al. in a Saudi population (Aljarbou et al. 2019). In a large Brazilian CBCT analysis of more than 9,000 roots, Oliveira et al. found limited or absent clearance between the distal roots of mandibular second molars and the IAC, whereas first molars generally provided greater available apical bone height (Oliveira et al. 2019). Choi et al. also demonstrated that mandibular second-molar roots frequently approached or contacted the lingual cortical plate and that this relationship varied according to vertical facial pattern (Choi et al. 2023). Similar findings have been reported in an Egyptian population (Sharaan et al. 2022). Beyond tooth-specific differences, Srivastava et al. documented considerable interindividual variation in the proximity of mandibular root apices to the IAC and adjacent cortical plates, supporting patient-specific three-dimensional evaluation rather than reliance on average anatomical measurements (Srivastava et al. 2022).

The interradicular septum may provide an important source of engagement when an implant is positioned within a multirooted molar socket (Sayed et al. 2021). Sayed et al. demonstrated that interradicular bone dimensions influence the feasibility of achieving primary stability during immediate molar implant placement (Sayed et al. 2021). Míchel-Baltazar et al. showed that septal width and root angulation vary considerably among mandibular first molars and may limit interradicular osteotomy preparation or necessitate alternative implant positioning (Míchel-Baltazar et al. 2023). Khoury et al. reported that septal width and planned implant position affect radiographic bone-to-implant contact (Khoury et al. 2023). Kong et al. similarly demonstrated that mandibular molar morphology influences prosthetically driven implant positioning and angulation (Kong et al. 2021). Ho et al. identified greater anatomical limitations at mandibular second-molar sites, including reduced interradicular bone, decreased vertical clearance above the IAC, and a higher prevalence of undercut ridge morphology (Ho et al. 2022). Virtual implant-placement studies have likewise shown that septal dimensions, implant position, and surrounding alveolar bone morphology should be assessed together (Chen et al. 2021).

CBCT studies conducted in Saudi Arabia have provided population-specific information regarding alveolar bone morphology and anatomical structures relevant to implant treatment (Sheerah et al. 2019). Sheerah et al. documented considerable variation in the facial alveolar bone plate of maxillary anterior teeth in a Saudi population (Sheerah et al. 2019). A CBCT investigation conducted at King Abdulaziz University similarly identified marked site-specific variation in the facial bone plate of maxillary anterior teeth (Othman et al. 2022). In the mandibular region, Othman et al. demonstrated that CBCT was more effective than panoramic radiography in detecting the anterior loop of the mental nerve (Othman et al. 2022). These studies, however, focused primarily on anterior teeth, facial bone dimensions, or mental nerve anatomy rather than providing a comprehensive assessment of mandibular first- and second-molar sites.

Previous CBCT investigations have evaluated overlapping aspects of mandibular molar anatomy, including cortical dimensions, interradicular bone, and proximity to the IAC; several have also combined anatomical assessment with virtual implant placement (Ebrahimi Dastgurdi et al. 2025, Shi et al. 2024, Khoury et al. 2023, Kong et al. 2021, Ho et al. 2022, Chen et al. 2021). The present study therefore does not introduce new anatomical parameters or establish implant-placement feasibility.

Its specific contribution is the combined assessment of six complementary anatomical parameters at all four mandibular first- and second-molar sites within the same 25 participants from a dental center in Jeddah, Saudi Arabia. This within-participant design permits paired comparisons of molar type and side while accounting for the related observations from each individual. The resulting measurements provide a descriptive anatomical comparison rather than a validated implant-planning protocol.

Accordingly, this retrospective cross-sectional CBCT study aimed to evaluate and compare buccal bone thickness, crestal bone width, crestal bone height loss, vertical bone height, root apex-to-IAC distance, and interradicular distance at mandibular molars #36, #37, #46, and #47, with potential relevance to preoperative anatomical assessment.

Materials and methods

Study design and setting

This retrospective cross-sectional morphometric study used a convenience sample of 45 CBCT scans retrieved from the database of Tam Dental Clinic, Jeddah, Saudi Arabia. The scans had been acquired for clinical purposes between January and December 2025 for diagnostic indications including orthodontic assessment, endodontic evaluation, third-molar surgery, and implant planning. Retrospective screening and data collection for the present study were conducted during January and February 2026. Because the study was retrospective and exploratory, no prospective recruitment or a priori sample-size calculation was performed. Of the 45 CBCT scans screened, 20 were excluded after application of the predefined eligibility criteria: 8 because of dental implants or previous root canal treatment involving the evaluated mandibular molars and 12 because of absence of one or more evaluated mandibular molars or periodontal involvement that prevented reliable measurement. No scans were excluded because of poor diagnostic image quality or an isotropic voxel size greater than 0.25 mm. The remaining 25 scans were included in the final analysis, with each scan contributing measurements from four mandibular molars, yielding 100 tooth sites.

This retrospective study was reviewed and approved by the Research Ethics Committee, Faculty of Dentistry, King Abdulaziz University, Jeddah, Saudi Arabia (Proposal No. 109-06-25; approval date: December 4, 2025). Ethics and consent details are provided in Section "Ethics approval and consent to participate"

Eligibility criteria

Inclusion criteria

• Adults aged 20–70 years.

• Presence of all four mandibular molars (#36, #37, #46, and #47).

Exclusion criteria

• Dental implants or previous root canal treatment involving the evaluated mandibular molars.

• Absence of one or more evaluated mandibular molars or periodontal involvement that prevented reliable measurement.

• Poor diagnostic image quality or an isotropic voxel size greater than 0.25 mm.

The final sample comprised 25 adults aged 22–55 years, all of whom had bilateral permanent mandibular first and second molars.

Examiner allocation, calibration, and measurement reliability

The 25 included scans were allocated between two trained periodontists (B.O. and L.A.) according to availability. B.O. evaluated 13 scans and L.A. evaluated 12 scans. For inter-examiner calibration, six included scans were independently re-evaluated by both examiners while each remained blinded to the other’s measurements.

These scans represented 24 mandibular molar sites and provided 24 paired observations for each parameter. Reliability was assessed separately for each parameter using a two-way random-effects intraclass correlation coefficient for absolute agreement and single measurements [ICC(2,1)]. Corresponding 95% confidence intervals were estimated using percentile bootstrapping with 50,000 paired resamples. ICC values were interpreted as poor (< 0.50), moderate (0.50–0.74), good (0.75–0.89), or excellent (≥ 0.90). For the primary analysis, the measurement recorded by the examiner originally assigned to each scan was retained; the additional calibration measurements were used exclusively for reliability assessment.

Radiographic assessment

CBCT datasets were reviewed using Planmeca Romexis software (Planmeca, Helsinki, Finland). The scans had an 80-mm field of view and were acquired at 120 kVp and 5 mAs, with an isotropic voxel size of 0.25 mm. For each mandibular molar, multiplanar CBCT images were reoriented to obtain a standardized section through the center of the tooth. The reference section was positioned through the tooth midline, bisecting the tooth into approximately equal halves. The image orientation was adjusted as required to clearly visualize the anatomical landmarks relevant to each parameter. Linear measurements were obtained using the software’s digital caliper tool and recorded in millimeters.

Image orientation and measurement definitions

Parameter A: Buccal bone thickness: the root was divided into cervical, middle, and apical thirds on the standardized tooth-oriented section. Buccal bone thickness was measured at the midpoint of the cervical third as the perpendicular distance between the external buccal cortical surface and the adjacent root surface (Fig. 1) (Heimes et al. 2021, Shi et al. 2024).

Fig. 1.

Measurement of buccal bone thickness (Parameter A). Representative tooth-oriented CBCT section demonstrating the perpendicular distance between the external buccal cortical surface and the adjacent root surface at the midpoint of the cervical root third

Parameter B: Crestal bone width: the buccolingual distance between the external surfaces of the buccal and lingual cortical plates at the alveolar crest on the standardized tooth-oriented section (Fig. 2) (Ebrahimi Dastgurdi et al. 2025, Kong et al. 2021, Chen et al. 2021).

Fig. 2.

Measurement of crestal bone width (Parameter B). Representative buccolingual CBCT cross-section demonstrating the buccolingual width of the alveolar ridge at the alveolar crest. The line indicates the distance between the external surfaces of the buccal and lingual cortical plates

Parameter C: Crestal bone height loss: the vertical distance from the buccal cementoenamel junction (CEJ), used as a fixed coronal reference point, to the buccal alveolar bone crest on the same standardized section (Fig. 3) (Fleiner et al. 2013).

Fig. 3.

Measurement of crestal bone height loss (Parameter C). Representative CBCT section demonstrating the vertical distance from the buccal cementoenamel junction to the buccal alveolar crest.27

Parameter D: Vertical bone height: the distance from the alveolar crest to the superior border of the IAC on the reoriented section that clearly demonstrated both landmarks (Fig. 4) (Ebrahimi Dastgurdi et al. 2025, Ho et al. 2022, Chen et al. 2021, Behfarnia et al. 2025).

Fig. 4.

Measurement of vertical bone height (Parameter D). Representative buccolingual CBCT cross-section demonstrating the distance from the alveolar crest to the superior border of the inferior alveolar canal

Parameter E: Root apex-to-IAC distance: the distance from the relevant root apices to the superior border of the IAC was assessed, and the root apex demonstrating the shortest distance to the IAC was used as the representative value for that tooth (Fig. 5) (Bürklein et al. 2015, Behfarnia et al. 2025).

Fig. 5.

Measurement of root apex-to-inferior alveolar canal distance (Parameter E). The root apex with the shortest distance to the superior border of the inferior alveolar canal was used as the representative value for each tooth

Parameter F: Interradicular distance: the interradicular region between the mesial and distal roots was divided into three approximately equal levels—coronal, middle, and apical. At each level, the horizontal distance between the opposing mesial and distal root surfaces was measured on the mesiodistal section. The three measurements were averaged to obtain a single representative interradicular-distance value for each tooth (Fig. 6) (Sayed et al. 2021).

Fig. 6.

Measurement of interradicular distance (Parameter F). Representative mesiodistal CBCT section. The interradicular region was divided into coronal, middle, and apical thirds; the horizontal distance between opposing mesial and distal root surfaces was measured at each level, and the three measurements were averaged

Image interpretability and recorded observations

All six parameters were successfully measured at all 100 included tooth sites. Scans in which image quality, restorative artifacts, periodontal involvement, or indistinct anatomical boundaries prevented reliable measurement were not eligible for inclusion under the predefined screening criteria; none of the included scans had an image-quality limitation that prevented measurement. When more than one potential measurement was available, the predefined anatomical rule for that parameter was applied consistently: the nearest root apex was retained for Parameter E, and the mean of the coronal, middle, and apical measurements was used for Parameter F.

Statistical analysis

Quantitative variables were summarized using means, standard deviations, and 95% confidence intervals. The normality of model residuals was assessed using the Shapiro–Wilk test and visual inspection of Q–Q plots.

Because each patient contributed measurements from four mandibular molars (#36, #37, #46, and #47), the observations were treated as related. Each anatomical parameter was analyzed using a 2 × 2 repeated-measures analysis of variance, with molar type (first versus second molar) and side (left versus right) entered as within-subject factors. The molar type × side interaction was also evaluated.

Significant effects were explored using paired comparisons. The six tooth-level comparisons conducted for each parameter were adjusted using the Holm procedure. Results are reported as mean differences with 95% confidence intervals, Holm-adjusted p-values, and paired Cohen’s d_z effect sizes. Overall effect sizes are reported as partial eta squared (partial η²).

For parameters in which the residual normality assumption was not met, complementary Wilcoxon signed-rank sensitivity analyses were performed. Statistical significance was set at p < 0.05. Statistical analyses were performed using Python version 3.13.5 and SciPy version 1.17.0. Microsoft Excel was used for data organization, archiving, and verification of the calculations.

Results

Participant characteristics

A total of 25 CBCT scans from 25 adults were included, contributing 100 mandibular molar sites: 25 sites each for #36, #37, #46, and #47. The sample comprised 13 women (52.0%) and 12 men (48.0%). The mean age was 35.16 ± 9.61 years (range, 22–55 years). Participant and dataset characteristics are presented in Table 1.

Table 1.

Participant and dataset characteristics

Characteristic	Value
Patients, n	25
Age, mean ± SD, years	35.16 ± 9.61
Age range, years	22–55
Women, n (%)	13 (52.0%)
Men, n (%)	12 (48.0%)
Included CBCT scans, n	25
Field of view, mm	80
Isotropic voxel size, mm	0.25
Evaluated molar sites, n	100
Tooth #36 sites, n	25
Tooth #37 sites, n	25
Tooth #46 sites, n	25
Tooth #47 sites, n	25

Values are presented as number, number (percentage), mean ± standard deviation, or range, as appropriate. CBCT, cone-beam computed tomography

Examiner reliability

Inter-examiner reliability was assessed using six CBCT scans representing 24 mandibular molar sites. ICC point estimates demonstrated excellent agreement for all six parameters and ranged from 0.907 to 0.988. Parameter-specific ICC values and confidence intervals are presented in Table 2.

Table 2.

Inter-examiner reliability of CBCT measurements

Parameter	Measurement	Paired observations	ICC(2,1)	95% CI	Interpretation
A	Buccal bone thickness	24	0.940	0.855–0.969	Excellent
B	Crestal bone width	24	0.971	0.923–0.988	Excellent
C	Crestal bone height loss	24	0.921	0.828–0.947	Excellent
D	Vertical bone height	24	0.985	0.958–0.995	Excellent
E	Root apex-to-IAC distance	24	0.988	0.968–0.995	Excellent
F	Interradicular distance	24	0.907	0.786–0.962	Excellent

ICC(2,1), two-way random-effects model, absolute agreement, single measurements. Confidence intervals were estimated using percentile bootstrapping with 50,000 paired resamples

Normality assessment

The residual normality assumption was met for buccal bone thickness, crestal bone width, vertical bone height, and interradicular distance. Residual normality was not met for crestal bone height loss (Shapiro–Wilk p = 0.003) and was borderline for root apex-to-IAC distance (p = 0.050). Complementary Wilcoxon signed-rank sensitivity analyses confirmed the conclusions of the repeated-measures analysis: crestal bone height loss remained nonsignificant, whereas root apex-to-IAC distance remained significantly greater at first-molar sites (p = 0.009). Descriptive measurements are presented in Table 3, overall repeated-measures effects in Table 4, and significant Holm-adjusted tooth-level comparisons in Table 5.

Table 3.

Descriptive statistics by tooth and anatomical parameter

Parameter	Tooth #36	Tooth #37	Tooth #46	Tooth #47
A. Buccal bone thickness	1.19 ± 0.41 (1.02–1.36)	1.57 ± 0.52 (1.36–1.79)	1.14 ± 0.33 (1.01–1.28)	1.47 ± 0.50 (1.26–1.67)
B. Crestal bone width	11.55 ± 1.56 (10.90–12.19)	12.23 ± 1.50 (11.61–12.85)	11.97 ± 1.92 (11.18–12.76)	12.86 ± 1.66 (12.17–13.54)
C. Crestal bone height loss	2.11 ± 0.64 (1.85–2.38)	2.11 ± 0.52 (1.90–2.33)	2.03 ± 0.48 (1.83–2.23)	2.24 ± 0.83 (1.90–2.58)
D. Vertical bone height	16.62 ± 2.96 (15.40–17.84)	15.18 ± 3.03 (13.93–16.43)	16.76 ± 2.87 (15.57–17.95)	14.75 ± 2.95 (13.53–15.96)
E. Root apex-to-IAC distance	5.43 ± 2.41 (4.44–6.43)	4.54 ± 2.41 (3.55–5.54)	4.90 ± 2.20 (3.99–5.81)	4.18 ± 2.02 (3.35–5.02)
F. Interradicular distance	2.77 ± 0.70 (2.48–3.06)	1.90 ± 0.50 (1.69–2.11)	2.70 ± 0.59 (2.45–2.94)	1.72 ± 0.56 (1.49–1.95)

Values are presented in millimeters as mean ± standard deviation (95% confidence interval). CBCT, cone-beam computed tomography, IAC, inferior alveolar canal

Table 4.

Repeated-measures analysis of molar type, side, and their interaction

Parameter	Shapiro–Wilk p	Molar-type difference¹ (95% CI)	F(1, 24)	p	Partial η²	Side difference² (95% CI)	F(1, 24)	p	Partial η²	Type × side p
A. Buccal bone thickness	0.062	0.352 (0.186–0.518)	19.173	< 0.001	0.444	−0.076 (− 0.218–0.066)	1.220	0.280	0.048	0.678
B. Crestal bone width	0.662	0.786 (0.367–1.205)	15.009	0.001	0.385	0.522 (0.106–0.938)	6.698	0.016	0.218	0.507
C. Crestal bone height loss	0.003	0.106 (− 0.104–0.316)	1.088	0.307	0.043	0.022 (− 0.216–0.260)	0.036	0.850	0.002	0.432
D. Vertical bone height	0.057	−1.724 (− 2.398 to − 1.050)	27.860	< 0.001	0.537	−0.144 (− 0.870–0.582)	0.168	0.686	0.007	0.288
E. Root apex-to-IAC distance	0.050	−0.802 (− 1.343 to − 0.261)	9.349	0.005	0.280	−0.446 (− 1.110–0.218)	1.921	0.179	0.074	0.700
F. Interradicular distance	0.851	−0.924 (− 1.193 to − 0.655)	50.195	< 0.001	0.677	−0.128 (− 0.288–0.032)	2.715	0.112	0.102	0.500

¹ Molar-type difference = mean of second molars minus mean of first molars. Positive values indicate larger measurements at second-molar sites; negative values indicate larger measurements at first-molar sites. ² Side difference = right minus left. The residual normality assumption was not met for crestal bone height loss and was borderline for root apex-to-IAC distance; complementary Wilcoxon signed-rank analyses confirmed the conclusions. IAC, inferior alveolar canal

Table 5.

Significant Holm-adjusted tooth-level comparisons

Parameter	Comparison	Mean difference, mm	95% CI	Holm-adjusted p	Cohen’s dz	Effect
Buccal bone thickness	#36 vs. #37	−0.380	−0.611 to − 0.149	0.012	−0.678	Moderate
	#37 vs. #46	0.428	0.242 to 0.614	< 0.001	0.948	Large
	#46 vs. #47	−0.324	−0.522 to − 0.126	0.012	−0.674	Moderate
Crestal bone width	#36 vs. #37	−0.684	−1.152 to − 0.216	0.024	−0.604	Moderate
	#36 vs. #47	−1.308	−1.896 to − 0.720	< 0.001	−0.917	Large
	#46 vs. #47	−0.888	−1.460 to − 0.316	0.019	−0.641	Moderate
Vertical bone height	#36 vs. #37	1.436	0.522 to 2.350	0.010	0.649	Moderate
	#36 vs. #47	1.868	0.746 to 2.990	0.009	0.687	Moderate
	#37 vs. #46	−1.580	−2.419 to − 0.741	0.004	−0.777	Moderate
	#46 vs. #47	2.012	1.192 to 2.832	< 0.001	1.013	Large
Root apex-to-IAC distance	#36 vs. #47	1.248	0.420 to 2.076	0.028	0.622	Moderate
Interradicular distance	#36 vs. #37	0.872	0.560 to 1.184	< 0.001	1.154	Large
	#36 vs. #47	1.052	0.708 to 1.396	< 0.001	1.263	Large
	#37 vs. #46	−0.796	−1.076 to − 0.516	< 0.001	−1.175	Large
	#46 vs. #47	0.976	0.665 to 1.287	< 0.001	1.295	Large

Mean differences were calculated as the first listed tooth minus the second listed tooth. Only comparisons with Holm-adjusted p < 0.05 are presented. Cohen’s dz represents the paired standardized mean difference. A negative value indicates that the second listed tooth had the larger measurement. The Holm procedure was applied separately within each anatomical parameter

Buccal bone thickness

Buccal bone thickness differed significantly by molar type, F(1, 24) = 19.173, p < 0.001, partial η² = 0.444. Second molars were, on average, 0.352 mm thicker than first molars (95% CI, 0.186–0.518 mm). No significant side effect or molar type × side interaction was detected. After Holm adjustment, #37 had greater buccal bone thickness than #36 and #46, and #47 was greater than #46 (Fig. 7).

Fig. 7.

Buccal bone thickness by mandibular tooth site. Individual observations are shown with the mean and 95% confidence interval. Brackets indicate statistically significant Holm-adjusted paired comparisons

Crestal bone width

Crestal bone width was significantly greater at second-molar sites, F(1, 24) = 15.009, p = 0.001, partial η² = 0.385. The mean difference between second and first molars was 0.786 mm (95% CI, 0.367–1.205 mm). A significant side effect was also identified, F(1, 24) = 6.698, p = 0.016, partial η² = 0.218; right-sided molars were 0.522 mm wider than left-sided molars (95% CI, 0.106–0.938 mm). No significant interaction was detected. After Holm adjustment, #37 was greater than #36, and #47 was greater than #36 and #46 (Fig. 8).

Fig. 8.

Crestal bone width by mandibular tooth site. Individual observations are shown with the mean and 95% confidence interval. Brackets indicate statistically significant Holm-adjusted paired comparisons

Crestal bone height loss

Crestal bone height loss did not differ significantly by molar type, F(1, 24) = 1.088, p = 0.307, partial η² = 0.043, or side, F(1, 24) = 0.036, p = 0.850, partial η² = 0.002. The interaction was also nonsignificant (p = 0.432), and no tooth-level comparison remained significant after Holm adjustment (Fig. 9).

Fig. 9.

Crestal bone height loss by mandibular tooth site. Individual observations are shown with the mean and 95% confidence interval. No tooth-level comparison remained significant after Holm adjustment

Vertical bone height

Vertical bone height was significantly greater at first-molar sites, F(1, 24) = 27.860, p < 0.001, partial η² = 0.537. First molars had, on average, 1.724 mm more vertical bone height than second molars (95% CI, 1.050–2.398 mm). Neither side nor the interaction was significant. Both first molars had greater vertical bone height than both second molars after Holm adjustment (Fig. 10).

Fig. 10.

Vertical bone height by mandibular tooth site. Individual observations are shown with the mean and 95% confidence interval. Brackets indicate statistically significant Holm-adjusted paired comparisons

Root apex-to-IAC distance

Root apex-to-IAC distance was significantly greater at first-molar sites, F(1, 24) = 9.349, p = 0.005, partial η² = 0.280. First molars demonstrated a mean distance 0.802 mm greater than second molars (95% CI, 0.261–1.343 mm). No significant side effect or interaction was identified. Among individual comparisons, only #36 versus #47 remained significant after Holm adjustment (mean difference, 1.248 mm; 95% CI, 0.420–2.076 mm; adjusted p = 0.028; Fig. 11).

Fig. 11.

Root apex-to-inferior alveolar canal distance by mandibular tooth site. Individual observations are shown with the mean and 95% confidence interval. The bracket indicates the significant Holm-adjusted comparison between #36 and #47 (Holm-adjusted p = 0.028). IAC, inferior alveolar canal

Interradicular distance

Interradicular distance showed the largest molar-type effect, F(1, 24) = 50.195, p < 0.001, partial η² = 0.677. First molars had an average interradicular distance 0.924 mm greater than second molars (95% CI, 0.655–1.193 mm). No significant side effect or interaction was observed. Both first molars had greater interradicular distances than both second molars after Holm adjustment (Fig. 12).

Fig. 12.

Interradicular distance by mandibular tooth site. Individual observations are shown with the mean and 95% confidence interval. Brackets indicate statistically significant Holm-adjusted comparisons between first- and second-molar sites

Discussion

This retrospective CBCT study demonstrated distinct anatomical patterns among mandibular first and second molars. Second molars showed greater overall buccal bone thickness and crestal bone width, whereas first molars demonstrated greater vertical bone height, root apex-to-IAC distance, and interradicular distance. Crestal bone height loss did not differ significantly by molar type or side. No molar type × side interaction was detected for any parameter in this sample. This does not establish bilateral equivalence or exclude smaller side-dependent differences. These findings provide morphometric information that may support site-specific preoperative assessment; however, they do not establish implant feasibility, primary stability, survival, or clinical success.

Statistical significance should be distinguished from clinical importance. The significant molar-type differences in buccal bone thickness (0.352 mm), crestal bone width (0.786 mm), root apex-to-IAC distance (0.802 mm), and interradicular distance (0.924 mm) were all smaller than 1 mm. These are group-level estimates, not validated thresholds for individual implant decisions. No decision thresholds, virtual implant placement, or clinical outcomes were evaluated; the clinical implications of these differences remain unestablished.

Buccal bone thickness and crestal bone width

Second molars demonstrated greater overall buccal bone thickness than first molars, with a large molar-type effect. At the tooth level, #37 was thicker than #36 and #46, while #47 was thicker than #46 after adjustment for multiple comparisons. Not every comparison between a first and second molar remained significant, so the overall molar-type effect should not be interpreted as a uniform difference across every tooth pair.

This finding is consistent with the systematic review by Heimes et al., which reported substantial site-related variation in buccal bone thickness and generally greater thickness in posterior than anterior regions (Heimes et al. 2021). Ebrahimi Dastgurdi et al. similarly found that cortical dimensions increased toward the second-molar region, although second molars simultaneously presented greater risks related to the IAC and lingual cortical plate (Ebrahimi Dastgurdi et al. 2025). Shi et al. reported that the buccal cortical plate surrounding mandibular first molars was relatively thin, particularly adjacent to the mesial root (Shi et al. 2024). The agreement with these recent studies suggests that differences in cortical and interradicular morphology may be reproducible across populations, although their magnitude remains population- and site-dependent.

The greater buccal thickness at second-molar sites may be related to progressive posterior widening of the mandibular body and alveolar process. Nevertheless, cortical thickness alone does not establish suitability for immediate implantation; root position, socket morphology, post-extraction remodeling, and the remaining facial bone envelope must also be considered. ^7,8,20

Crestal bone width was also greater at second molars, with a large molar-type effect. A significant side effect was observed, with right-sided molars showing greater crestal width than left-sided molars; however, no molar type × side interaction was detected in this sample. This side-related finding should be interpreted cautiously because it may reflect anatomical asymmetry or the relatively small sample rather than a stable population-level pattern. Greater crestal width describes a larger buccolingual bone envelope, but this study did not test implant-diameter selection or a planned implant position.

Vertical bone height and root apex-to-IAC distance

First molars demonstrated greater vertical bone height than second molars, with a large molar-type effect. Both first molars had greater vertical height than both second molars in the adjusted comparisons. This pattern may reflect the posterior course of the mandibular canal and its increasingly close relationship with molar roots and the alveolar crest toward the second-molar region.

Oliveira et al. reported that mandibular first molars generally provided more available apical bone than second molars, while distal roots of second molars frequently showed limited clearance from the mandibular canal (Oliveira et al. 2019). Bürklein et al. likewise found that second-molar root apices were closer to the canal than first-molar roots (Bürklein et al. 2015). Comparable findings have been reported in Saudi, Egyptian, and other populations (Aljarbou et al. 2019, Sharaan et al. 2022, Srivastava et al. 2022). Choi et al. additionally reported frequent proximity of second-molar roots to the lingual cortical plate (Choi et al. 2023). A recent systematic review and meta-analysis by Behfarnia et al. confirmed reduced root apex-to-IAC clearance and a greater prevalence of undercut ridge morphology at mandibular second-molar sites (Behfarnia et al. 2025).

Root apex-to-IAC distance was greater overall at first-molar sites, although only #36 versus #47 remained significant after Holm adjustment. The complementary nonparametric analysis confirmed the molar-type difference, supporting the stability of the finding despite a minor departure from residual normality.

The greater vertical height and root apex-to-IAC distance at first molars may provide more available apical bone during planning. However, implant length and drilling depth cannot be selected from mean measurements alone. The complete three-dimensional anatomy, required safety margin from the IAC, socket configuration, implant design, and prosthetically planned position must be assessed for each patient.

Interradicular distance and septal anatomy

Interradicular distance showed the largest molar-type effect. Both first molars had greater interradicular distances than both second molars, with no statistically significant difference detected between the two first molars or between the two second molars. These results describe a molar-type difference in this sample, without establishing equivalence between sides.

The greater interradicular distance at first molars may be related to greater root divergence and broader furcation morphology. In contrast, second molars more frequently have convergent or partially fused roots, which may reduce septal width. Sayed et al. reported that interradicular bone dimensions are important when evaluating potential primary stability during immediate molar implant placement (Sayed et al. 2021). Míchel-Baltazar et al. demonstrated substantial variation in root angulation and septal width among mandibular first molars (Míchel-Baltazar et al. 2023). Khoury et al. found that septal width and implant position influenced radiographic bone-to-implant contact (Khoury et al. 2023).

The Smith–Tarnow classification relates molar socket morphology to the bone available to engage an immediate implant: type A has sufficient septal bone to surround the coronal implant, type B provides septal engagement without complete circumferential containment, and type C requires engagement of the socket periphery because septal support is insufficient (Smith and Tarnow 2013). These categories depend on socket anatomy in relation to a planned implant, not on a single linear root-separation measurement.

Subsequent approaches further emphasize the combination of septal anatomy, implant position, and socket-wall integrity. Khoury et al. considered septal width in relation to implant position and radiographic bone-to-implant contact (Khoury et al. 2023). Ekhlasmandkermani et al. proposed a framework incorporating the condition of both the interradicular septum and the buccal bone plate (Ekhlasmandkermani et al. 2025). Ghishan et al. demonstrated variation in root-trunk, root-concavity, and furcation morphology using direct digital analysis of extracted posterior teeth; their anatomical findings should not be interpreted as validation of an immediate-implant socket classification (Ghishan et al. 2024).

In the present study, interradicular distance was measured at three standardized levels within the interradicular region of dentate sites, and the mean of these measurements was used as the representative value for each tooth. The study did not assess a post-extraction socket, evaluate a virtual implant within it, or measure primary stability. Therefore, the observed distances cannot assign a Smith–Tarnow socket type, establish circumferential implant engagement, or predict primary stability.

Recent CBCT and virtual implant-planning studies have similarly emphasized that interradicular width, root divergence, implant position, and surrounding alveolar morphology should be assessed together rather than treating septal width as an isolated measurement (Ebrahimi Dastgurdi et al. 2025, Shi et al. 2024, Khoury et al. 2023, Kong et al. 2021, Ho et al. 2022, Chen et al. 2021). Although first-molar sites had greater measured interradicular distances, the present study did not test interradicular implant engagement; primary stability cannot be predicted from septal width alone because it also depends on bone quality, socket integrity, implant geometry, osteotomy preparation, insertion torque, and the extent of apical or lateral engagement.

Crestal bone height loss

Crestal bone height loss did not differ significantly by molar type or side, and no adjusted tooth-level comparison was significant. The complementary nonparametric analysis likewise detected no difference; these findings do not establish equivalence of crestal bone levels among the four sites.

Although no between-tooth difference was identified, crestal bone height remains clinically relevant. Pre-existing periodontal bone loss may affect the residual socket walls, available anchorage, implant platform position, and the potential need for regenerative procedures. The absence of a significant group difference should therefore not be interpreted as evidence that the parameter can be disregarded during individual planning.

Potential anatomical relevance of the evaluated parameters

The six CBCT-derived measurements describe complementary aspects of site-specific anatomy relevant to preoperative assessment (Ebrahimi Dastgurdi et al. 2025, Shi et al. 2024, Sayed et al. 2021, Kong et al. 2021, Ho et al. 2022, Chen et al. 2021, Fleiner et al. 2013). Buccal bone thickness and crestal bone width characterize the facial and lingual bone envelope,^6–8,18,20 while crestal bone height loss describes existing periodontal support (Fleiner et al. 2013). The present study did not establish how these measurements should determine implant diameter, position, or regenerative treatment.

Vertical bone height and root apex-to-IAC distance describe vertical and apical relationships to the IAC. Interradicular distance describes the separation of the mesial and distal roots at the measured level (Sayed et al. 2021). These anatomical observations may inform preoperative assessment, but their ability to determine osteotomy depth, drilling angulation, or a safe and feasible implant position was not tested.

Collectively, these parameters provide descriptive information about the bone envelope, interradicular anatomy, and proximity to the IAC. They have not been validated here as a decision rule for implant dimensions or osteotomy position. Primary stability also depends on factors not assessed in this study, including bone quality, socket integrity, implant design, osteotomy preparation, and insertion torque (Witoonkitvanich et al. 2025, Sayed et al. 2021, Khoury et al. 2023, Chen et al. 2021).

Recent digital-guidance studies have emphasized the importance of translating CBCT-derived planning accurately to the surgical site. Geng et al. reported reduced positional and angular deviations with dynamic navigation during immediate posterior mandibular implant placement, while Mahmoud et al. demonstrated improved positioning accuracy with computer-guided surgery compared with freehand placement (Geng et al. 2024, Mahmoud et al. 2024). These findings support the clinical value of comprehensive CBCT assessment but do not imply that guided surgery can compensate for unfavorable anatomy or inadequate case selection.

Relationship to clinical outcome evidence

Systematic reviews have reported favorable survival outcomes for immediate molar implants when appropriate case selection and surgical protocols are applied (Atieh et al. 2010, Ragucci et al. 2020). More recent prospective evidence has demonstrated high primary stability for computer-guided immediate implants placed in maxillary and mandibular molar sockets (Witoonkitvanich et al. 2025). A one-year CBCT-based study reported comparable survival and clinically acceptable peri-implant bone remodeling after immediate and delayed loading in the posterior mandible (Kaur et al. 2025). Favorable five-year outcomes have also been reported for immediate implants placed in mandibular molar sockets with chronic apical periodontitis after careful debridement and infection control (Yang et al. 2023).

These clinical studies provide context for the potential use of immediate molar implant protocols. The present investigation, however, assessed only preoperative radiographic anatomy and did not evaluate extraction sockets, infection status, implant placement, insertion torque, stability, healing, marginal bone remodeling, or survival. The findings should therefore be interpreted as potentially informative preoperative anatomical measurements rather than validation of implant-planning decisions or clinical effectiveness.

Limitations

Several limitations should be considered. First, the retrospective design limited control over patient selection, the original indications for CBCT imaging, and the availability of potentially relevant clinical variables. The use of a convenience sample may also have introduced selection bias.

Second, the study included only 25 patients. Although the scans contributed 100 mandibular molar sites and the analysis accounted for repeated measurements within each patient, the relatively small patient-level sample may have reduced the precision of the estimates and the ability to detect smaller anatomical effects.

Third, all scans were obtained from a single dental center in Jeddah, Saudi Arabia. The findings may therefore not represent the broader Saudi population or populations with different ethnic, anatomical, or geographic characteristics, limiting external validity and generalizability.

The sample was also restricted to selected dentate sites with all four mandibular molars present and anatomy that met the measurement eligibility criteria. Molars requiring extraction may have periodontal destruction, compromised socket walls or septa, or periapical disease that differs from these selected sites. The findings therefore cannot be transferred directly to immediate implant sockets, and the requirement for four retained molars may underrepresent clinically compromised dentitions.

Fourth, the study assessed linear CBCT-derived dimensions only. Bone density and trabecular quality were not evaluated, and three-dimensional socket morphology, cortical integrity after extraction, infection status, and post-extraction remodeling were not assessed.

Finally, this was not a clinical implant outcome study. No virtual implant placement or clinical implant placement was performed as part of the investigation, and insertion torque, implant stability quotient, surgical complications, marginal bone changes, implant survival, and treatment success were not evaluated. Consequently, the radiographic differences cannot be directly correlated with feasibility, primary stability, survival, or clinical success.

Future directions

Future multicenter studies should include larger and more diverse populations and incorporate volumetric and bone-density assessments. Prospective clinical investigations are also needed to correlate preoperative CBCT measurements with intraoperative socket morphology, implant dimensions, insertion torque, primary stability, surgical complications, peri-implant remodeling, and long-term implant survival.

Conclusion

This retrospective CBCT study identified distinct anatomical differences between mandibular first and second molars. Second molars demonstrated greater overall buccal bone thickness and crestal bone width, whereas first molars showed greater vertical bone height, root apex-to-IAC distance, and interradicular distance. Crestal bone height loss did not differ significantly by molar type or side.

These CBCT-derived measurements may be informative for preoperative anatomical assessment of the bone envelope, interradicular anatomy, and proximity to the IAC in mandibular molar regions. However, the findings were obtained from selected dentate sites and have not been validated for selecting implant dimensions, determining osteotomy positions, assigning extraction-socket types, or predicting primary stability or clinical outcomes. Studies incorporating virtual implant placement and prospective clinical validation are needed to establish their decision-making value.

Author contributions

B.O. was solely responsible for the conception and design of the study, data acquisition and interpretation, statistical analysis and verification, preparation of figures and tables, drafting of the manuscript, critical revision of the intellectual content, and approval of the final manuscript.

Funding

The authors received no financial support for the research, authorship, or publication of this article.

Data availability

The datasets generated and analyzed during the current study are not publicly available because of institutional and patient privacy considerations but are available from the corresponding author upon reasonable request and subject to institutional approval.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the Research Ethics Committee, Faculty of Dentistry, King Abdulaziz University, Jeddah, Saudi Arabia (Proposal No. 109-06-25; approval date: December 4, 2025). The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The study involved the retrospective analysis of anonymized CBCT images previously acquired for clinical purposes, with no additional imaging, intervention, or direct patient involvement. Individual written informed consent was considered not applicable because only retrospective anonymized imaging data were analyzed, as documented in the ethics application reviewed and approved by the REC. Written institutional permission to access and use the CBCT data for research purposes was also obtained from Tam Dental Clinic, Jeddah, Saudi Arabia.

Consent for publication

Not applicable. The manuscript contains no identifiable personal data, patient photographs, or other information requiring individual consent for publication.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work, the authors used ChatGPT (OpenAI) to assist with language editing, manuscript organization, and the development and verification of statistical-analysis code. Following use of this tool, the authors reviewed and verified the manuscript content, references, statistical outputs, and interpretations and take full responsibility for the content of the article.

Competing interests

The authors declare no competing interests.