Application of the logistic regression model for the analysis of accuracy related influencing factors on HPLC measurement of preservatives
10.3969/j.issn.1006-2483.2021.03.012
- VernacularTitle:logistic回归模型分析HPLC测防腐剂准确度相关因素
- Author:
Bin WANG
1
;
Yao LI
1
;
Ying ZHANG
1
;
Jiuhong LI
1
;
Xin ZHANG
1
;
Xinyue CHEN
1
;
Li MAO
1
;
Kui LIU
2
Author Information
1. Department of Public Health, Chengdu Medical College, Chengdu 610500, China
2. Sichuan Center for Disease Control and Prevention, Chengdu 610041, China
- Publication Type:Journal Article
- Keywords:
HPLC;
Preservation;
Recovery;
Logistic regression analysis
- From:
Journal of Public Health and Preventive Medicine
2021;32(3):50-53
- CountryChina
- Language:Chinese
-
Abstract:
Objective To employ Logistic regression modeling to analyze the related factors influencing the accuracy of the high performance liquid chromatography (HPLC) determination of preservatives in beverages. Methods The HPLC separation was performed on a Zorbax Eclipse Plus C18 column with methanol-ammonium acetate solution as mobile phase. The external standard method was used to determine 5 beverage preservatives. The influencing factors on the measurement accuracy were statistically evaluated with univariate and multivariate analysis. Results Univariate analysis showed that the recovery rate of the added standard in the determination of coffee beverage preservatives by HPLC was affected by the pretreatment method, and the difference was statistically significant (P<0.05). Multivariate analysis showed that the main influencing factors on the accuracy of determination of sorbic acid was the pretreatment method (OR=5.406, P<0.05), while the sample type was a protective factor (OR=0.134, P<0.05). For the determination of benzoic acid, the main factor influencing the accuracy was the sample type (OR=1.112, P<0.05), while the pretreatment method was a protective factor (OR=0.447, P<0.05). Conclusion Logistic regression analysis can identify risk factors for the accuracy of the determination, and provide statistical modeling support for the experimental optimization.
