Dietary inadequacy and nutrition-related non-communicable diseases (N-NCDs) represent two main issues for the whole society, urgently requesting solutions from researchers, policy-makers, and other stakeholders involved in the health and food system. Food by-products and wastes (FBPW) represent a global problem of increasing severity, widely recognized as an important unsustainability hotspot, with high socio-economic and environmental costs. Yet, recycling and up-cycling of FBPW to produce functional foods could represent a solution to dietary inadequacy and risk of N-NCDs onset. Bioprocessing of FBPW with selected microorganisms appears to be a relatively cheap strategy to yield molecules (or rather molecules mixtures) that may be used to fortify/enrich food, as well as to formulate dietary supplements. This review, conjugating human health and sustainability in relation to food, describes the state-of-the-art of the use of yeasts, molds, and lactic acid bacteria for producing value-added compounds from FBPW. Challenges related to FBPW bioprocessing prior to their use in food regard will be also discussed: (i) loss of product functionality upon scale-up of recovery process; (ii) finding logistic solutions to the intrinsic perishability of the majority of FBPW; (iii) inserting up-cycling of FBPW in an appropriate legislative framework; (iv) increasing consumer acceptability of food and dietary supplements derived from FBPW.

An untargeted method using direct analysis in real time and high resolution mass spectrometry (DART-HRMS) combined to multivariate statistical analysis was developed for the discrimination of two monofloral (chestnut and acacia) honeys for their geographical origins—i.e., Italy and Portugal for chestnut honey and Italy and China for acacia honey. Principal Component Analysis, used as an unsupervised approach, showed samples of clusterization for chestnut honey samples, while overlapping regions were observed for acacia honeys. Three supervised statistical approaches, such as Principal Components—Linear Discriminant Analysis, Partial Least Squares—Discriminant Analysis and k-nearest neighbors, were tested on the dataset gathered and relevant performances were compared. All tested statistical approaches provided comparable prediction abilities in cross-validation and external validation with mean values falling between 89.2–98.4% for chestnut and between 85.8–95.0% for acacia honey. The results obtained herein indicate the feasibility of the DART-HRMS approach in combination with chemometrics for the rapid authentication of honey’s geographical origin.