Comparison of radial immunodiffusion, turbidimetric immunoassay, and Brix refractometry for determining bovine colostrum quality

Determining the concentration of IgG in colostrum is critical for assessment of colostrum quality. On-farm use of a Brix refractometer to estimate colostrum IgG concentration is widespread, whereas radial immunodiffusion (RID) is the laboratory reference method. Turbidimetric immunoassay (TIA) might offer an alternative method to quantify IgG in colostrum, but the agreement with RID, as well as critical thresholds to determine high-quality colostrum, remain uncertain. The objective of this study was to determine the level of agreement between RID, Brix %, and TIA for evaluation of colostrum quality. Composite colostrum samples (n = 58) from Holstein cows were evaluated using a digital Brix refractometer at the time of collection and stored at −20°C until analysis. The concentration of IgG was determined using RID and TIA. Data were analyzed using Passing-Bablok regression and Bland-Altman plots. Critical thresholds for TIA and Brix measurements to identify colostrum with an IgG concentration ≥50 and ≥100 g/L based on the reference method were determined using logistic regression and receiver operating characteristic curves. Results revealed that both TIA (Kendall's τ = 0.91) and Brix % (τ = 0.78) had a strong correlation with RID. Passing-Bablok regression identified a systematic (6.91 [4.33 to 8.98] g/L) and proportional (0.69 [0.67 to 0.72] g/L) bias between RID and TIA. The optimum thresholds to identify samples with an IgG concentration ≥50 and ≥100 g/L were 40.6 g/L (area under the curve [AUC]: 1.0; sensitivity (Se): 100; specificity [Sp]: 100) and 85.8 g/L (AUC: 0.99; Se: 96.6; Sp: 96.6) for TIA and 18.4% (AUC: 1.0; Se: 100; Sp: 100) and 25.8% (AUC: 0.99; Se: 82.8; Sp: 93.1) for Brix %, respectively. Using the identified thresholds, our results show that both Brix and TIA were highly accurate for identifying high-quality colostrum, but because of a proportional bias, direct comparison of IgG concentration results obtained by RID and TIA are cautioned.