Sex determination and stature estimation using logistic and linear regression models: A population-specific study of tibia and Fibula in Northeastern Thais

Background

Sex determination and stature estimation from skeletal remains are basic components of forensic anthropological analysis, particularly when fragmentary remains limit conventional approaches. This study aimed to establish the population-specific standards for Northeastern Thais using tibial and fibular measurements.

Materials and methods

The study examined 200 identified adult skeletons (100 males, 100 females). The nine osteometric parameters (six tibial and three fibular) were measured bilaterally. All specimens were systematically documented for age, sex, and stature.

Results

For sex determination, the results showed that univariate logistic regression models achieved the accuracy rates up to 87.50 %, with maximum proximal epiphyseal breadth of the tibia demonstrating highest discriminatory power. Multivariate approaches could significantly enhance the accuracy to 91.30 % for left tibial parameters, with age-related variation revealing optimal results in the 45–59 age group. Fibular measurements showed lower accuracy but still maintained statistical significance, particularly for fibular weight. For stature estimation, the maximum tibial length served as the most reliable univariate predictor (SEE = 6.66 cm, R = 0.572), while multivariate equations incorporating complementary parameters from both tibia and fibula improved the accuracy. Age-stratified analysis demonstrated diminished accuracy in older cohorts, especially for fibular measurements.

Conclusion

The findings align with the previous studies while reflecting distinct patterns unique to Northeastern Thai populations with their specific genetic ancestry. These population-specific equations can enhance the biological profiling capabilities in forensic investigations when the conventional methods are inapplicable due to fragmentary remains.