Comparison of spectrophotometric and HPLC methods for determination of carotenoids in foods

This report is aimed at intra-laboratory and inter-laboratory comparison of the results obtained during spectrophotometric and HPLC analyses of lycopene, β-carotene and total carotenoids in tomato products and yellow maize flours/grits. Extensive statistical analyses are performed in order to identify the main sources of uncertainties which may occur when using: (i) different techniques/methods/approaches in the same/different laboratories, in various food samples, and (ii) to indicate the facts/conditions under which the biases between results may remain unidentified after applying statistical testing. Our data points to the inertness of t-test to detect significance of differences, particularly at low R values: in general, the higher correlation coefficient, the higher is sensitivity of statistical testing, especially of the paired t-test. Therefore, simple deviation of relationship line slope from unity could be used as additional evaluation parameter. This adds to reliable and objective quality assurance of foods in regard to carotenoids.

► Statistical comparison of spectrophotometric and HPLC data for carotenoids in food. ► Lycopene, beta carotene, total carotenoids in tomato products and maize flours/grits. ► Intra-laboratory/inter-laboratory comparison of results for carotenoids in food. ► Inertness of statistics (R and unpaired t-test); uncertainty of results evaluation. ► Line slope deviation from unity as average bias estimate.