Abstract
BACKGROUND: Malnutrition is a common geriatric syndrome that is closely associated with adverse clinical outcomes and poses significant harm to older adults. Early assessment of nutritional status plays a crucial role in preventing and intervening in cases of malnutrition. However, there is currently a lack of measurable methods and biomarkers to evaluate malnutrition in older adults accurately. The aim of this study is to investigate the independent correlation between serum levels of amino acids and malnutrition in older adults, and to identify effective metabolomics biomarkers that can aid in the early detection of geriatric malnutrition. METHODS: A total of 254 geriatric medical examination participants from Beijing Hospital were included in the study, consisting of 182 individuals with normal nutritional status (Normal group) and 72 patients at risk of malnutrition or already malnourished (MN group). Malnutrition was assessed using the Mini-Nutritional Assessment Short-Form (MNA-SF). Demographic data were collected, and muscle-related and lipid indexes were determined. Serum amino acid concentrations were measured using isotope dilution liquid chromatography-tandem mass spectrometry (LC-MS/MS). The correlation between serum amino acid levels and malnutrition was analyzed using non-parametric tests, partial correlation analysis, linear regression, and logistic regression. RESULTS: The geriatric MN group exhibited significantly lower serum aromatic amino acid levels (P < 0.05) compared to the normal group. A positive correlation was observed between serum aromatic amino acid levels and the MNA-SF score (P = 0.002), as well as with known biomarkers of malnutrition such as body mass index (BMI) (P < 0.001) and hemoglobin (HGB) (P = 0.005). Multivariable logistic or linear regression analyses showed that aromatic amino acid levels were negatively correlated with MN and positively correlated with the MNA-SF score, after adjusting for some confounding factors, such as age, gender, BMI, smoking status, history of dyslipidemia, diabetes mellitus and frailty. Stratified analyses revealed that these trends were more pronounced in individuals without a history of frailty compared to those with a history of frailty, and there was an interaction between aromatic amino acid levels and frailty history (P = 0.004). CONCLUSION: Our study suggests that serum aromatic amino acids are independently associated with malnutrition in older adults. These results have important implications for identifying potential biomarkers to predict geriatric malnutrition or monitor its progression and severity, as malnutrition can result in poor clinical outcomes.