Heterocyclic aromatic amines in roasted chicken: Formation and prediction based on heating temperature and time. (30th March 2023)

• Record Type: Journal Article
• Authors: Yao, Guangming; Zhou, Yajun; Li, Zongping; Ma, Qingshu
• Is Part Of: Food chemistry. Volume 405:Part A(2023)
• Publication Date: 2023-03-30
• Subjects: Heterocyclic aromatic amine -- Roasted chicken -- Temperature -- Time -- Precursor -- Artificial neural network; Food -- Analysis -- Periodicals; Food -- Composition -- Periodicals; 664
• DOI: 10.1016/j.foodchem.2022.134822
• Languages: English
• ISSNs: 0308-8146

Abstract

Highlights: Roasted chicken contained IQ, MeIQ, MeIQx, 4, 8-DiMeIQx, PhIP, harman, and noharman. Transportation of precursors may contribute greatly to the high HAA contents in skin. PCA shows HAAs change significantly with the increase of temperature and time. A HAA prediction model was built using backpropagation-artificial neural network. This study facilitates the development of intelligent real-time HAA detection. Abstract: The effects of chicken roasting temperature and time on the production of heterocyclic aromatic amines (HAAs) were investigated and an HAA prediction model based on heating conditions was established. Generally, the HAA content was significantly affected by the heating conditions in the roast chicken. Transportation of precursors from meat to skin, exposure of skin to high temperatures, and fat oxidation in the skin may result in higher HAAs than meat. Principal component analysis (PCA) showed that the effect of relatively high temperatures and long roasting times on HAAs was stronger than that of lower temperatures and shorter roasting times. In the prediction of HAA production, all regression correlation coefficient (R) values were close to one. The errors of 15 samples of experimental and predictive data were close to zero. Based on the results, backpropagation-artificial neural network (BP-ANN) has a high potential for predicting the production of HAAs under heating conditions.