Materials and Methods: A model of ATR-FTIR-PLSR was developed using ATR-FTIR spectra of mixed aflatoxin standards in 100% acetonitrile (112 samples) and 75% methanol (112 samples), validated by testing its prediction on 125 feed/food samples spiked with variable concentrations of aflatoxins, and applied to screen 660 samples of commercial chicken feeds and food grains from Nigerian and Malaysian markets for total aflatoxins, for which the dietary exposure risks to aflatoxins (DERA) and associated hepatocellular carcinoma (HCC) risks were evaluated for both countries.

Results: The ATR-FTIR-PLSR model demonstrated excellent prediction power [R 2 = 99.59%, p = 0.001, root mean square error of calibration (RMSEC) = 1.69, RMSE p = 1.98, bias = -0.26], sensitivity (limit of quantitation and limit of the method < 5.0 ng/gm), precision (coefficient of variation = 0.97-1.72), and accuracy (% recovery of 88%-106%) in all the spiked samples. The model's prediction was statistically reliable (R 2 = 99.8%, p < 0.05) when compared with a high-performance liquid chromatography method. Levels of aflatoxins in the commercial samples signify high DERA (0.92-138.2 ng of aflatoxins/kg BW/day) and HCC risk (1.07%-159.91% of HCC/100,000 people/year) in the exposed populations.

MATERIALS AND METHODS: To obtain the optimal mobile phase, samples were extracted with methanol/water (3:1) + 5% sodium chloride and partitioned using several solvent systems using preparative TLC. Camag TLC scanner 3 was used to scan the TLC plates at 366 nm and quantify them using JustTLC software. The method was tested for linearity, specificity, accuracy, precision, sensitivity, and robustness in accordance with ICH recommendations, and then utilized to screen 132 Nigerian poultry/food samples for total aflatoxins (TAFs).

RESULTS: The best separation of aflatoxins was achieved using acetonitrile and dichloromethane (3:17) mobile phase over an average run time of 45 min, resulting in linear calibration curves (R2 > 0.99) in the concentration range limit of quantitation (LoQ) to 50 ng/spot with a limit of detection of <2.0 ng/g and a LoQ of <4.0 ng/gm for all aflatoxins in all spiked samples. When the proposed TLC method was compared to an optimized high-performance liquid chromatography method, an excellent linear regression was obtained (R2 > 95%). Seventy seven (58.33%) of the 132 samples examined were positive for aflatoxins, with mean values ranging from 3.57 ± 2.55 to 37.31 ± 34.06 ng/gm for aflatoxin B1 and 6.67 ± 0.00 to 38.02 ± 31.52 ng/gm for TAFs, respectively.