Sexual Dimorphism in Canine Teeth for Forensic Sex Determination: A Population-Specific Study from Northern Ghana

Background : Sexual dimorphism in dental morphology is fundamental to forensic identification, and population-specific variations are critical for accurate sex determination. Despite its importance in medicolegal investigations, comprehensive data on dental sexual dimorphism remain limited in West African populations. Objective : This study evaluated the forensic utility of maxillary and mandibular canine dimensions for sex determination in the Northern Ghanaian population and assessed the practical limitations of this approach. Methods : A cross-sectional study examined 212 participants aged 18-25 years [93 males (43.9%), 119 females (56.1%)] from the Tamale region of Northern Ghana (May-September 2022). Mesiodistal canine widths and inter-canine distances were measured using standardized anthropometric protocols with an inter-rater reliability assessment. Statistical analysis employed descriptive statistics, binary logistic regression, and diagnostic accuracy measures, including ROC curve analysis, with significance set at p<0.05. Results : Mandibular canines demonstrated modest sexual dimorphism, with males showing significantly larger dimensions: right mandibular canine width (7.26±0.67mm vs. 6.88±0.51mm, p<0.001), left mandibular canine width (7.15±0.60mm vs. 6.87±0.39mm, p<0.001), and inter-canine distance (30.88±2.74mm vs. 29.32±1.98mm, p<0.0001). Traditional forensic canine indices showed an accuracy equivalent to chance (49.5-51.9%, 95% CI: 43.2-58.6%). Binary logistic regression achieved marginally improved performance, with the left mandibular canine index reaching a maximum accuracy of 58.0% (95% CI: 51.1-64.7%, AUC=0.618). Conclusions : Despite statistically significant sexual dimorphism in Northern Ghanaian canines, forensic utility is severely limited by accuracy rates barely exceeding chance (58% vs. 50%). The 8% improvement over random classification falls substantially below the 80-90% threshold required for reliable forensic applications. Canine-based sex determination lacks practical forensic value as a standalone method in this population