Evaluation and predictive modelling of okara (poly)phenol metabolite production via lactic acid fermentation

Okara, a soy-processing by-product rich in bioactive compounds like isoflavones, represents a promising substrate for valorisation. This study explored how lactic acid bacteria (LAB) and Bacillus subtilis, used either in mono- or in co-culture, can enhance the phenolic composition of okara through controlled fermentation. The design of experiment was applied for testing four strains across different inoculum levels, fermentation temperatures, and times, and phenolic profiles were quantified using UHPLC-QqQ-MS/MS. LAB fermentation significantly improved the phenolic content and profile of okara, particularly using Pediococcus pentosaceus and Lactobacillus paracasei under extended fermentation at 25 °C, resulting in up to 38% increase in total phenolics. Fermentation promoted the bioconversion of isoflavone glycosides into aglycones through β-glucosidase activity. Phenyllactic acid derivatives were detected, although no other low molecular weight phenolics were observed. Thanks to the application of predictive modelling the variables most affecting phenolic outcomes were identified and optimal fermentation conditions could be defined. When co-cultures were used, the fermentation process strongly influenced bioactive compound profile, highlighting the need for tailored microbial combinations. Overall, these findings demonstrated an efficient biotechnological approach for converting okara into a functional food ingredient with enhanced potential health-promoting properties, supporting its industrial valorisation within circular bioeconomy frameworks.