ESPEN 2022 Abstract Submission

Topic: Nutritional assessment

Abstract Submission Identifier: ESPEN22-ABS-1593

IS DIABETES ASSOCIATED WITH A SPECIFIC PHENOTYPE OF MALNUTRITION AND MORTALITY? AN OBSERVATIONAL STUDY IN HOSPITALIZED PATIENTS MANAGED BY THE NUTRITIONAL SUPPORT TEAM OF AN UNIVERSITY HOSPITAL

M. Barrallier^*, ¹, E. Rageul ¹, C. Nesseler ¹, S. Martin ², R. Thibault ^{1, 3}

¹Service Endocrinologie-Diabétologie-Nutrition, ²Centre d’Investigation clinique, INSERM 1414, CHU Rennes, ³INSERM, INRAE, Nutrition Metabolisms and Cancer, NuMeCan, Univ Rennes, Rennes, France

Rationale: Hospital malnutrition is common, but its phenotype in diabetic patients is not well known. Aims: in hospitalized patients followed by the Nutrition Support Team (NST): i) to describe the phenotype of malnutrition to determine whether diabetes was a risk factor for severe malnutrition (primary); ii) to determine the clinical outcomes (secondary); and iii) the predictive factors of mortality (secondary).

Methods: We conducted a retrospective observational monocentric study at our University Hospital. Patients followed by the NST were included between November 2019 and May 2021. All patients were performed a body composition assessment by bioimpedance analysis, and, if possible, an assessment of muscle function by handgrip strength. Malnutrition was diagnosed according the Global Leadership Initiative on Malnutrition criteria. The etiologic criterion was the hospitalization for an acute care. Statistics: multivariable analysis using stepwise logistic regression. 

Results: 231 patients were included: 193 (83.5%) were malnourished, and 141 (73%) severely, 47 (20%) patients (women, 74.5%, type 2 diabetes, 68%) were diabetic. Malnutrition prevalence did not differ between diabetic and non-diabetic patients (39 (83.0%) vs 154 (83.7%), p=0.91), and the same was for severe malnutrition (24 (61.5%) vs. 117 (76.0%), p=0.07). Malnutrition phenotype was significantly different in diabetic patients: higher BMI (mean±SD, 26.3±6.4 vs. 22.8±6.3, p=0.001), higher proportion of patients with decreased muscle strength (measured in 83% of patients) (64.3 vs. 45.6%, p=0.03), while their fat-free mass index (FFMI) was paradoxically higher (15.5±4.1 vs 17.9±3.4 kg/m², p<0.0001). At 1 month, diabetic patients reported more abdominal infections (20 (42.6%) vs 50 (27.2%), p=0.04) and bacteremia (18 (38.3%) vs 34 (18.5%), p=0.004), and a trends toward higher mortality (9 (19.1%) vs. 17 (9.2%), P=0.06). Refeeding syndrome, pressure ulcers, length of hospital stay and 1-month hospital readmission were not different. No independent predictive factor of mortality was found, but diabetes was the last remaining variable in the multivariable analysis (odds ratio=2.33 [95% confidence interval, 0.96; 5.62], P=0.06).

Conclusion: Malnutrition is frequent in hospitalized diabetic patients managed by the NST. The malnutrition phenotype seems to be different: lower muscle strength and higher FFMI. Diabetic patients were at higher risk for infections. Our data suggest that thresholds for muscle mass and strength should be refined to diagnose malnutrition in diabetic patients.

Disclosure of Interest: None Declared

Keywords: bioimpedance analysis, Diabetes Mellitus, malnutrition, hospitalization