Elevated GDF-15 levels may indicate malnutrition in chronic compensated heart failure with or without diabetes mellitus

Abstract

Abstract Background/Introduction Malnutrition is common in chronic heart failure (HF) and is associated with adverse outcomes. Elevated NT-proBNP may indicate increased cardiac stress, whereas growth differentiation factor 15 (GDF-15) is increased in response to tissue injury, inflammation and in diabetes mellitus (DM). GDF-15 also negatively regulates appetite, body weight and skeletal muscle homeostasis. Purpose We aim to assess the nutritional status of HF patients with or without DM, and determine if GDF-15 is associated with malnutrition. Methods We did a cross-sectional analysis on chronic HF patients in a prospective cohort study, forecAsting Heart Failure decompensation (AHF). Chronic compensated HF was defined as a state following hospital discharge for >30 days, or where there had not been hospitalisation for 30 days preceding phenotyping. Patients with a history of HF decompensation were included. Geriatric Nutritional Risk Index (GNRI), including body mass index (BMI) and serum albumin, was used to assess and categorise nutritional status. Patients were classified as being “malnourished” or “well-nourished” based on GNRI scores of <92 and ≥92, respectively. NT-proBNP and GDF-15 levels were measured and echocardiography was done. The study was approved by institutional review board and complied with Declaration of Helsinki. Results Among 73 patients (mean age 66 y, 58.9% male), 32.7% exhibited malnutrition as determined by GNRI. There was no difference in age, sex ratio and NYHA functional class between the malnourished and the well-nourished. BMI (22.1 vs 25.6, P<0.001) and levels of GDF-15 (7216.0 vs 2469.0, P<0.001) and NT-proBNP (7158.0 vs 1321.0, P=0.01) were different between the malnourished and well-nourished. Both GDF-15 and NT-proBNP were inversely correlated with GNRI (GDF-15, r=−0.39, P<0.001; NT-proBNP, r=−0.46, P<0.0001). As expected, DM HF patients had significantly higher levels of GDF-15 than non-DM HF patients. In subgroup analysis of DM HF patients, GDF-15 and NT-proBNP levels were higher in the malnourished than well-nourished (median log-[GDF-15], 9.15 vs 8.10, P<0.01; median log-[NT-proBNP], 9.44 vs 7.00, P<0.01). Whereas in non-DM HF, NT-proBNP levels were not different between the malnourished and well-nourished. Age- and sex-adjusted logistic regression model determined that per log increment in GDF-15 level there was a 5.8-fold increased risk of malnutrition (OR 5.81 [2.43–17.62], P<0.001), and for every log increment in NT-proBNP level there was a 1.8-fold increase in the risk of malnutrition (OR 1.82 [1.23–2.88], P<0.01). Conclusion(s) Malnutrition is common in chronic compensated HF patients with a history of hospitalisation, and is associated with elevated NT-proBNP and GDF-15 levels. Elevated GDF-15 levels independently predicted poor nutritional status, and with NT-proBNP may indicate a particularly high-risk HF subgroup Funding Acknowledgement Type of funding source: Public Institution(s). Main funding source(s): 1) Health and Medical Research Fund, Food and Health Bureau, 2) Direct Grant, Faculty of Medicine, CUHK