The association between malnutrition and dehydration in older adults admitted to a geriatric unit: An observational study

Abstract

Background and aim: There is an overlap between the risk factors causing low intake of water and low intake of nutrients, respectively. This study aims to explore the agreement between the assessment of malnutrition and the outcome of low-intake dehydration in a population of older hospitalized patients.

Methods: Patients ≥65 years old and hospitalized at the geriatric hospital ward were screened for eligibility within 96 h of admission. Dehydration was assessed with the calculated serum osmolarity ≥295 mmol/L (1.86 × (Na⁺ + K⁺) + 1.15 × glucose + urea + 14), and (risk of) malnutrition was assessed with NRS-2002 ≥ 3 points, MNA-SF ≤ 7 points, MNA-LF < 17, MUST ≥ 2 points, and GLIM after screening with NRS-2002 and MNA-LF. Follow-up data regarding exercise rehabilitation, readmissions, and mortality was collected 30 days after discharge. Statistics used were the Chi-squared test, Fishers-exact test, and Wilcoxon signed rank test.

Results: A total of 114 patients (57% females) were included. Median age 85.5 (IQR 80; 89.25) years. A total of 49 (43%) were dehydrated. Fewer females were dehydrated (F: 42.9% vs. M: 67.7%, p = 0.013). The patients with osmolarity ≥295 mmol/L had a higher median weight (68.3 (IQR 58.5; 78.4) vs. 62 (IQR 51.8; 72.1), p = 0.021) and mid-up-arm circumference (27 (IQR 26; 30) vs. 25.5 (IQR 22.9; 28.3), p = 0.004). No significant difference was found in the prevalence of malnutrition between those with or without dehydration (NRS-2002; 70% vs. 81%, p = 0.174; MNA-SF: 23.1 vs. 23.2%, p = 1.0; MNA-LF: 37.1 vs. 30.2%, p = 0.644; MUST: 24.5 vs. 33.8%, p = 0.308; GLIM after screening with NRS-2002: 84.4 vs. 74.5%, p = 0.405, GLIM after screening with MNA-LF: 74.1 vs. 75.6%, p = 0.438). Kappa values varied around 0 and reflected low agreement. There were no differences in the follow-up data, between those who were normohydrated and those who were dehydrated.

Conclusion: We found low agreement between the assessment of malnutrition and low-intake dehydration in a population of older hospitalized patients. All geriatric patients should therefore be assessed for both conditions.

Keywords: GLIM; Mortality; Nutritional risk; Re-admission; Serum osmolarity.