Preoperative Prediction of Total Small Bowel Length Using Anthropometric Measurements: Development and Internal Validation of a Model in 327 Adult Autopsies

Background/Objectives: Total small bowel length (TSBL) exhibits considerable interindividual variability and may influence outcomes following bariatric and metabolic surgery. This study aimed to evaluate the relationship between anthropometric characteristics and TSBL and to develop and internally validate an anthropometry-based prediction model. Methods: In this prospective observational study, 327 adult autopsy cases were included. Height, weight, body mass index, waist circumference, and abdominal fat thickness were recorded using a standardized protocol. TSBL was measured from the ligament of Treitz to the ileocecal junction. Correlation analyses, multivariable linear regression, and prediction model development were performed. Internal validation was conducted using 1000 bootstrap resamples. Model performance was additionally assessed using prediction error, calibration, and agreement between predicted and observed TSBL. Results: Mean TSBL was 612.4 ± 108.7 cm. Height showed the strongest correlation with TSBL (r = 0.612, p < 0.001), followed by weight (r = 0.428, p < 0.001) and waist circumference (r = 0.341, p < 0.001). In multivariable analysis, height (standardized β = 0.447, p < 0.001), weight (standardized β = 0.185, p = 0.006), and waist circumference (standardized β = 0.099, p = 0.029) remained independently associated with TSBL. The final prediction model yielded R2 = 0.456 and adjusted R2 = 0.451, with a root mean square error of 80.1 cm and mean absolute error of 62.8 cm. Bootstrap validation yielded an optimism-corrected R2 of 0.444 and a calibration slope of 0.97. Predicted TSBL was within ±10% of the observed value in 56.0% of cases. Conclusions: Height was the strongest independent predictor of TSBL, while weight and waist circumference provided additional predictive information. The proposed anthropometry-based model showed moderate predictive performance but retained substantial individual prediction error. It should therefore be considered an exploratory tool that may provide supplementary preoperative information rather than a substitute for direct bowel measurement. External validation in living and independent populations is required before clinical implementation.