Prediction of body weight and ethnicity using anthropomorphic measurements of the hand in two different populations

This study aimed to evaluate the usefulness of hand and finger measurements for prredicting body weight and distinguishing between Egyptians and Saudis. Also, to design regression models to predict body weight from hand and finger measurements. A descriptive cross-sectional study was conducted among 160 adult volunteers from two population groups (80 Egyptians and 80 Saudis), aged 20 to 27. The measurements taken were body weight and anthropometric measurements of the right hand, which comprised the hand length, handbreadth, palm length, and total and phalangeal finger length. Univariate and multivariate linear regression and ROC curve analysis were performed in the statistical analyses. Egyptians had no statistically significantly bigger hands than the Saudis. Egyptian and Saudi male participants had statistically significantly larger hand measurements than their corresponding female. Various hand and finger measurements were strong predictors of body weight and could differentiate Egyptian and Saudi ethnicities with sufficient accuracy. The highest precision was observed in Saudi females, where the little finger measurement yielded the most accurate model (R2 = 0.817, SEE = 3.15 kg). Egyptian males showed a moderate predictive capacity (R2 = 0.634, SEE = 7.14 kg), heavily influenced by handbreadth, while Egyptian females relied on index and middle finger phalanges to achieve a model with R2 = 0.525. Saudi males presented the least precise multivariate results with an R2 = 0.553 and the highest margin of error (SEE = 14.453 kg). The integration of multiple linear regression improved estimation accuracy compared to univariate models, which generally exhibited higher error ranges between 9.13 kg and 11.33 kg. The index proximal phalangeal length (AUC = 0.715) has the most potent discriminatory ability to distinguish between Saudi and Egyptian males (sensitivity 85.0%, cut-off ≤ 2.6). For females, Handbreadth was the best ethnic discriminator with an Area under the Curve (AUC) of 0.818 (sensitivity 75.0%, cut-off ≤ 7.8). Anthropometric measurements of the hand and fingers could be used to predict weight and ethnicity in both Egyptian and Saudi populations. This may draw attention to the anticipated practical medico-legal identification.