Evaluation of quality consistency of Astragali radix by multi-dimensional profiles combined with chemometric analysis

Abstract

Astragali radix (AR) is not only a traditional Chinese medicine, but also a good tonic that is often consumed by people and has good disease prevention and health care effects. Due to the complex composition of AR and the interaction of different chemical components, it presents challenges in quality control. In this study, a gradient elution tower plate theory suitable for gradient elution was firstly established, followed by HPLC analysis combined with quantitative analysis of multi-components by single marker (QAMS) and multi-markers assay by monolinear method (MAML) for accurate content determination of the indicator components (calycosin-7-O-β-D-glucoside and formononetin) in AR. FTIR and DSC are also used to collect the fingerprints of AR and quantize them to obtain the structural information inside the substance. The established fingerprints were evaluated qualitatively and quantitatively using the systematic quantitative fingerprint method (SQFM) and chemometrics, and the characteristic parameters of the three fingerprints were investigated in relation to the macro-quantitative similarity (P_m) using the Pearson correlation analysis. Finally, the coefficient of variation weighted mean method (CVWMM) was used to synthesize the evaluation results and establish the HPLC–FTIR–DSC quality control system, which is more objective and reasonable. The results showed that the calculated results of QAMS and MAML were basically consistent with those of an external standard method. P_m had the function of representing the overall fingerprint information of the samples for quantitative evaluation. SQFM combined with CVWMM successfully differentiated the quality grades of 35 batches of AR samples. The combination of the three analytical methods resulted in the effective improvement of the quality evaluation of AR compared with a single method in terms of accuracy and completeness. This study provides a novel and comprehensive strategy for quality consistency evaluation of AR.