Existing and emerging in vitro and in vivo methods for protein and amino acid assessment in foods: a systematic review of methodological advances, validation, and predictive performance

Protein and amino acid assessment is central to nutrition, food formulation and regulatory evaluation, particularly as novel and sustainable protein sources expand. This systematic review synthesizes 2016–2026 evidence on methods for protein quantity/quality, amino acid composition and digestibility, with emphasis on methodological comparison, validation and predictive performance. The evidence spans nitrogen-based methods, amino acid hydrolysis and chromatography, static and dynamic in vitro digestion, in vivo ileal digestibility, DIAAS and PDCAAS, spectroscopy, peptidomics, microfluidics and artificial intelligence (AI). Amino acid methods generally reported recoveries of approximately 90–105% and relative standard deviations (RSDs) below about 3–8%, whereas spectroscopy/AI studies commonly reported R2 values of approximately 0.85–0.99. Selected in vitro validation studies reported correlations above r = 0.8, but these were matrix-specific and did not establish agreement or absence of systematic bias; reported paired differences of approximately 5–15% between some in vitro and in vivo digestibility estimates likewise varied by matrix and protocol. These findings indicate that analytical accuracy, agreement with a reference method, biological validity and regulatory validity are distinct evidence domains. Persistent sources of uncertainty include matrix effects, non-protein nitrogen, incomplete or differential amino acid recovery, digestion-model assumptions, inter-laboratory variability, limited external validation of AI models, and incomplete translation of in vitro findings to human physiological utilization. The synthesis therefore supports a staged evidence framework in which chemical measurement feeds amino acid recovery, digestibility estimation and DIAAS calculation, followed by validation against physiological outcomes before methods are used for definitive nutritional or regulatory decisions.