D102-03 Examining the Relationship Between Ulnar Length and Standing Height in an Ethnically Diverse Population

Rationale Standing height is required for reliable pulmonary function test (PFT) predicted values, yet in patients with musculoskeletal and neuromuscular diseases, standing height often cannot be reliably measured, leading to errors in predicted values which can cause underestimation of pulmonary disease evaluation and monitoring. Current height estimation models, such as the Gauld Model, utilize ulnar length as a substitute measurement; however, this model was derived from a relatively racially homogenous population in Australia with notable limited sample size of Black (n = 3) and Hispanic (n = 13) patients, calling to question generalizability to diverse clinical settings. This study examined the performance of the Gauld Model in a racially heterogeneous cohort representative of the United States. Methods A single center study of patients presenting for PFTs in Hartford, Connecticut between 5-19 years of age, excluded for known musculoskeletal deformity or neuromuscular disease. Patients’ age, sex, race, ethnicity, height, weight, and right ulnar length were collected using data collection software (RedCap). Predicted height was determined using Gauld’s Model. Comparison across racial groups and sexes was performed using Pearson correlation and ANOVA. Results The sample population included 584 patients (21 Asian, 101 Black, 148 Hispanic, 314 White; 281 Female, 303 Male). Correlation (r), between our samples’ predicted height and actual height using the Gauld model, was 0.80, lower than Gauld’s correlation of 0.96. The mean error (M) between actual and predicted height per Gauld’s model was -4.0 cm with standard deviation (SD) of 8.8 cm.Between races no significant difference in correlation was found. However, mean error was significant (p=.012) between racial groups using the Gauld Model. Specifically, mean error was smallest in White patients (M=-2.9) and significantly increased in Asian (M= -7.0, p=.041), Black (M= -5.1, p=.034), and Hispanic (M= -5.2, p=.012) patients. Between sexes no significant difference in correlation was found. However, mean error was significantly larger (p=.030) in females (M= -5.2) than males (M= -3.7). No significant interaction between race and sex was found. Concluson The race-gender neutral Gauld model currently overestimates standing height in our patient population with significant mean error among the non-white and female patients. These findings call into question the race neutrality of Gauld’s predictive model in diverse clinical settings. We aim to develop a model that accurately predicts standing height from ulnar length measurements in our diverse population.