Association of serum albumin redox state with dietary protein intake and frailty in older adults: A cross-sectional study.
Journal:
Clinical nutrition ESPEN
Published Date:
Jul 3, 2026
Abstract
BACKGROUND AND AIMS: Biomarkers of protein nutritional status are essential for understanding the relationship between diet and frailty. The fraction of mercaptoalbumin in total serum albumin (f(HMA)) is associated with protein deficiency and physical function in older adults. However, whether f(HMA) explains the relationship between protein intake and frailty remains unclear. We aimed to determine whether f(HMA) reflects frailty severity and clarify the relationship between protein intake and nutritional biomarkers in older adults. METHODS: This cross-sectional study analyzed data from 377 adults aged ≥65 years. Participants were classified as robust, prefrailty, or frailty based on the revised Japanese version of the Cardiovascular Health Study criteria. Dietary protein intake was assessed using a brief-type self-administered diet history questionnaire. Protein nutritional status was evaluated using serum albumin, f(HMA), transthyretin, total protein, and branched-chain amino acids levels. Relationships between protein intake and biomarkers were analyzed using multivariate linear regression while associations with frailty severity were examined using ordinal logistic regression. Mediation analysis assessed whether f(HMA) mediated the effect of protein intake on frailty status (prefrailty-or-worse vs robust). In an exploratory analysis, machine learning models were developed to discriminate frailty using a random forest classifier. RESULTS: The prevalence of prefrailty and frailty was 52.5% and 14.7%, respectively. Protein intake was negatively associated with frailty severity (P = 0.041). Only f(HMA) was positively associated with protein intake (P = 0.005) and negatively associated with frailty or prefrailty (P < 0.001 and P = 0.008). f(HMA) was independently associated with protein intake and frailty status and partially statistically mediated their relationship (P = 0.009). The classification model incorporating f(HMA) outperformed traditional biomarker-based models, achieving an area under the receiver operating characteristic curve of 0.80. CONCLUSIONS: f(HMA) may serve as a useful biomarker linking protein nutrition to frailty in older adults. Further studies are required to confirm the predictive role of f(HMA) in frailty progression.
Authors
Keywords
No keywords available for this article.