Quantification of Total Sugars in Kombucha Using Attenuated Total Reflectance–Fourier Transform Infrared (ATR-FTIR) Spectroscopy and Partial Least Squares (PLS) Regression: A Rapid Alternative to Colorimetric Methods

The determination of total sugar content in kombucha is essential for monitoring the fermentation process, ensuring product safety, assessing nutritional value, and complying with labeling requirements. Additionally, it allows the prediction of ethanol and organic acid formation, which impact the sensory quality and preservation of the beverage. This study developed and validated a rapid, sustainable, and non-destructive method for quantifying total sugars in commercial Brazilian kombuchas, using mid-infrared spectroscopy by attenuated total reflectance with Fourier transform infrared (ATR-FTIR) spectroscopy combined with partial least squares regression (PLS). Spectral analysis was performed in the range of 4000 to 650 cm−1, and reference values were obtained using the spectrophotometric 3,5-dinitrosalicylic acid (DNS) method. Ninety-six commercial samples representing 30 Brazilian brands were evaluated, with sugar levels ranging from 20 to 84 g/L. The PLS model demonstrated excellent predictive performance, with calibration and prediction root mean square errors (RMSEC and RMSEP, respectively) of 2.8 and 2.7 g/L, respectively. Validation parameters such as linearity, recovery, precision, bias, and ratio of performance to deviation (RPD) confirmed the method's robustness and reliability for application in kombucha quality control. This work fills a gap in the application of vibrational spectroscopy to complex fermented matrices and provides original data on the composition of kombuchas in Brazil, highlighting the potential of ATR-FTIR as an effective analytical tool aligned with the demands of the food industry and regulatory agencies.