Screening analysis of garlic-oil capsules by infrared spectroscopy and chemometrics ☆

The search for a healthier life has led to an increase in the consumption of natural products with certain functional and/or nutritional properties while avoiding the collateral and harmful effects of medicines. In this context, garlic oil, in the form of gelatinous capsules, contains various substances considered beneficial to health, to reduce cholesterol and blood pressure, in the treatment of influenza and diabetes and to prevent the development of tumors and cardiovascular diseases. Currently, garlic oil is classified by ANVISA (National Health Surveillance Agency) as a “new food”, denomination given to food or substances without historical consumption in the country, or foods with substances already consumed and that may be added or used at much higher levels than currently observed in food. While there is no established law ensuring safety and quality of garlic oil commercialized in capsules, the development of simple and rapid methods of analysis to verify the authenticity of these products is of great importance. In this paper, we propose a screening analysis method to do this, using spectroscopy in the mid-infrared region and supervised pattern recognition methods, aimed at classifying garlic oil capsules by concentration levels. For this purpose, spectra of 136 samples from 11 brands of commercial capsules of and 123 solutions of garlic oil in soybean oil (0.00%, 0.25%, 0.50%, 0.75%, 1.00%, 1.25%, 1.50% and 1.75% (m/m)) were obtained using an attenuated total reflectance accessory. The spectra obtained were subjected to different preprocessing (Standard Normal Variate - SNV, Multiplicative Scatter Correction – MSC, and derivation by Savistzky Golay filters - SG) and uninformative spectral regions were eliminated. The supervised pattern recognition methods Partial Least Squares Discriminant Analysis (PLS-DA) and Linear Discriminant Analysis (LDA) with variable selection by the Successive Algorithm Projections (SPA) were used to build models capable for classifying commercial capsules into three classes: class 1 (low dosage) from 0.00% to 0.50%, class 2 (medium dosage, usual) from 0.75% to 1.25% and class 3 (high dosage) from 1.50% to 1.75%. The PLS-DA provided, in the validation stage, a correct classification rate of 81.3% against 95.9% of the SPA-LDA method. In the SPA-LDA classification, only the 924cm−1 variable was selected. This is where a band from the deformation of the vinyl groups of the sulfides and vinyl dithiins, main constituents of garlic oil, was observed. For the 136 capsules tested a rate of 88.2% agreement with the information contained on their labels was obtained. The high number of capsules classified in the wrong classes suggests that their labels had presented incorrect information about their concentration levels. The screening methodology proposed in this work is a simple, fast and effective tool that can be useful to investigate the content of garlic oil gelatinous capsules.