Independent validation of sex estimation equations using ulnar dimensions and weight in a northeastern Thai population

Sexual dimorphism in human skeletal remains is crucial for forensic identification and archaeological studies. This study develops and validates sex estimation equations using ulnar dimensions and weight in a northeastern Thai population. The study examined 600 ulnae (300 male, 300 female) from the Khon Kaen University skeletal collection, using 400 for equation development and 200 for independent validation. Maximum length, midshaft circumference, weight, and the weight-to-length ratio were assessed. Significant differences between sexes were found across all variables (p<0.05). Weight demonstrated the highest single-variable accuracy (88.50%), while combining circumference and weight achieved 90.80% accuracy. Validation confirmed equation stability, with no significant difference between the training and validation groups (90.75% vs. 90.50%, p=0.77). The equations showed excellent discriminatory ability (AUC=0.959), high specificity for female identification (95.00%), and good sensitivity for male identification (86.00%). The positive likelihood ratio of 17.20 further confirmed the robustness of the method. This research contributes to forensic anthropology by validating accurate sex estimation methods for the northeastern Thai population while demonstrating the effectiveness of bone weight as a sex determination variable. • Ulna weight: highest single-variable sex estimation accuracy (88.5%). • Circumference and weight combination achieves maximum accuracy (90.8%). • Validation confirms equation stability between training and test groups. • Statistical metrics (LR+=17.20, AUC=0.959) show excellent discriminatory power. • Large sample (600 ulnae) with exceptional measurement reliability (ICC>0.950).