Reference interval for serum amyloid a in apparently healthy donkeys measured with a donkey-specific ELISA

Abstract

Introduction Serum amyloid A (SAA) is a major acute phase protein used as an inflammatory biomarker in equine medicine. However, no universally accepted assay-specific reference interval (RI) exists for donkeys, and available literature remains limited, with only few studies using different donkey breeds and immunoassay platforms. This study aimed to establish an assay- and population-specific RI for serum SAA in apparently healthy donkeys using a species-specific double sandwich ELISA. Methods Serum SAA concentrations were measured in 176 apparently healthy donkeys. RI estimation followed ASVCP/CLSI recommendations. The primary RI was calculated nonparametrically as the central 95% interval from all reference individuals, and two-sided 90% confidence intervals (CIs) for the reference limits were estimated nonparametrically. Data were inspected using Tukey’s interquartile fences; however, high-end observations were not excluded from the primary RI analysis. Exploratory analyses were performed to assess the need for partitioning by sex, age, or breed. Results The nonparametric RI for serum SAA was 2.91–42.85 ng/mL. The 90% CI for the lower reference limit was 0.00–4.07 ng/mL, and the 90% CI for the upper reference limit was 35.49–67.10 ng/mL. Exploratory analyses did not support partitioning by sex, age, or breed. The relatively broad CI for the upper reference limit indicated imprecision, likely reflecting the influence of a minority of high-end observations. Discussion The low SAA concentrations observed in this study are consistent with findings reported in horses. Differences between this and previous donkey studies may reflect variation in assay platform, sample size, management conditions, physiological state, or population characteristics. Because SAA is a rapidly responsive acute phase protein, the proposed interval should be interpreted as an assay- and population-specific RI rather than as a disease-classification decision limit. Clinical interpretation should therefore always consider the individual donkey’s clinical context.