Poor appetite and growth differentiation factor-15 as predictors of insufficient energy and protein intake during and after hospitalization in older adults with acute medical illness: Exploratory analysis of a randomized controlled trial

BACKGROUND & AIMS: Poor appetite is a key contributor to malnutrition in older adults, partly due to reduced dietary intake. While nutritional deficits are well recognized, the biological mechanisms driving poor appetite remain incompletely understood. Growth Differentiation Factor-15 (GDF-15) and the Simplified Nutritional Appetite Questionnaire (SNAQ) may help identify patients at risk of insufficient intake. This study examines the association between GDF-15 and insufficient energy and protein intake and evaluates the predictive performance of GDF-15 and SNAQ - individually and in combination - to identify insufficient energy and protein intake (<75 % and <100 % of estimated requirements, respectively) in acutely admitted older adults.

METHODS: This exploratory study included 130 older adults (≥65 years) with or at risk of malnutrition, admitted for acute medical illness and assessed at baseline, and 8 and 16 weeks post-discharge (FW8 and FW16). GDF-15 plasma concentrations were measured from blood samples, SNAQ scores from validated questionnaire, and energy and protein intake were evaluated through 3-day dietary records. Associations were analyzed using regression models and predictive performance was evaluated using Receiver Operating Characteristic analysis.

RESULTS: A doubling of GDF-15 showed a 6.73 % (-13.98-0.51) and 5.07 % (-12.18-2.05) lower baseline energy and protein intake, and results decreased after discharge. Using predefined cut-offs of ≥1500 pg/mL for GDF-15 and ≤14 for SNAQ, the highest positive predictive values (PPVs) were 86 (95 % CI: 0.83-0.89) and 90 (0.79-1.00) at baseline and FW8 for insufficient protein intake, respectively, with corresponding sensitivities of 84 (0.75-0.92) and 60 (0.44-0.74). For estimated cut-offs of 2095 pg/mL for GDF-15 at baseline and 18.5 for SNAQ at FW8, the highest PPVs were 89 (0.84-0.94) and 87 (0.86-0.87), with sensitivities of 73 (0.62-0.84) and 99 (0.96-1.00), both for insufficient protein intake, respectively. Combining GDF-15 and SNAQ yielded improved PPVs at the expense of reduced sensitivity in most models. All results were non-significant.

CONCLUSION: Higher GDF-15 levels showed a non-significant trend toward lower baseline energy and protein intake. GDF-15 and SNAQ offer limited ability to identify insufficient energy and protein intake, based on predictive performance. These exploratory findings should therefore be interpreted cautiously. Larger studies are needed to validate and further refine the use of GDF-15 and SNAQ in clinical settings.