Sex and stature prediction from upper limb measurements: a forensic anthropological analysis in a Jordanian population

Forensic anthropology plays a key role in the identification process, especially in the estimation of important biological parameters like sex and stature, which are cornerstones in criminal investigations and disaster victim identification. While numerous studies have explored these parameters using lower and upper limb bones in various populations, data from Jordanians remain scarce. This study evaluates the predictive value of upper limb measurements for sex estimation and stature estimation in a Jordanian population, with the goal of establishing the first population-specific reference standards to strengthen forensic anthropological practice in Jordan. A sample of 200 right-handed Jordanian participants was included. Several upper limb measurements, including total arm length, upper arm length, forearm length, ulnar length, breadth of wrist, and length of hand, were recorded alongside height at the Faculty of Medicine, Mutah University. Correlation and regression analyses were applied to derive sex-specific stature estimation models, while discriminant function analysis was used to assess the potential of these variables for sex estimation. Participant ages ranged from 20 to 25 years old. All upper limb dimensions were significantly larger in males (p < 0.05). In males, ulnar length (r = 0.55) and upper arm length (r = 0.52) correlated most strongly with height, while in females, hand length (r = 0.68) and total arm length (r = 0.65) were the best predictors. The lowest standard error of estimate (SEE) for single-measurement prediction was obtained using ulnar length in males (5.65 cm) and hand length in females (4.49 cm). Multiple regression further improved prediction accuracy, reducing the SEE to 3.97 cm in females when using hand length, total arm length, and forearm length, and to 5.14 cm in males when using ulnar length, upper arm length, and total arm length. Ulnar length and wrist breadth were the best indicators for sex classification, with an accuracy of 84–85%. This cross-sectional study provides the first reference dataset for sex and stature estimation from upper limb measurements in a Jordanian population. Findings suggest the need for sex-specific, population-specific regression models, particularly in contexts where only upper limb remains are available. Future research should involve larger, more heterogeneous samples and strengthen the forensic applicability of the results by validating these equations on skeletal remains.