Background: Analysis of copper and zinc in serum is commonly performed using atomic absorption spectrometry (AAS); however, these methods are often not readily available in smaller laboratories. Randox colorimetric assays for copper and zinc in serum were evaluated on the Thermo Electron Data Pro analyser against flame AAS methods.
Methods: Copper and zinc were measured in 48 serum samples using the Randox colorimetric copper (CU2340) and zinc (ZN2341) assays on the Data Pro analyser and the results compared with those from a Varian Spectra 880 atomic absorption spectrometer. A smaller set of samples (n = 15) were also analysed colorimetrically for zinc on the Roche Cobas Mira.
Results: Linear regression analyses of Bland and Altman plots from the Data Pro - AAS comparison gave the following results for copper: correlation r = 0.6669 (P < 0.01), slope = -0.2499 (P < 0.01), intercept = 3.219 (P < 0.01). For zinc, results were as follows: correlation r = 0.1976, slope = 0.1807, intercept = -1.922. For the smaller set of samples, the Cobas Mira - AAS comparison for zinc gave correlation r = 0.4379, slope = 0.5294, intercept = -4.074. The results indicated significant systematic and fixed bias between the colorimetric copper and the AAS method.
Conclusion: Performances in comparison to AAS methods indicated the colorimetric methods, as used, are unsuitable for the accurate determination of copper and zinc in human serum.