Breaking the challenge of polyherbal quality control using UV and HPLC fingerprints combined with multivariate...

Traditional herbal medicines are mostly composed of more than one herb which act synergistically; hence, there is high demand for proper quality control methods to ensure the consistent quality of polyherbal formulations.

Proposing a simple, reliable, and efficient method for the qualitative and quantitative analysis of a polyherbal product using multivariate analysis of ultraviolet-visible (UV-Vis) spectroscopy or HPLC-PDA data.

An antiobesity formula consisting of equal proportions of Trachyspermum ammi, Cuminum cyminum, and Origanum majorana was prepared as well as spiked with one of each herb simultaneously, representing accepted and unaccepted samples. Full factorial design (2^k ) was applied to study the effect of temperature, sonication, and stirring time for extraction optimisation. The HPLC and UV spectral fingerprints were separately subjected to multivariate analysis. The soft independent modelling of class analogy (SIMCA) and partial least squares (PLS) models were developed to segregate the accepted from the unaccepted samples and to predict the herbal composition in addition to the thymol content in each sample.

The SIMCA^uv and SIMCA^hplc models showed correct discrimination between the accepted and unaccepted samples with excellent selectivity and sensitivity. The PLS^uv , PLS^hplc , and PLS^thym models showed excellent linearity and accuracy with R²  > 0.98, slope close to 1, intercept close to 0, low root mean square error of calibration (RMSEC), and root mean square error of prediction (RMSEP) (close to 0). On validation, the PLS models correctly predicted the quantity of the three herbs and thymol content with ±5% accuracy.

This study demonstrates the reliability and efficiency of HPLC and UV spectroscopy coupled with multivariate statistical analysis (MVA) for ensuring the consistency of polyherbal preparations.