Effects of Oral Nutritional Supplements in Nursing Home Residents

Although oral nutritional supplements (ONS) are known to be effective to treat malnutrition in older persons, evidence from nursing home populations including demented residents is rare, especially with regard to functionality and well-being. A known barrier for ONS use among older persons is the volume that needs to be consumed, resulting in low compliance and thus reduced effectiveness.

Thus, this study aimed to investigate the effects of a new, low volume, energy- and nutrient-dense ONS on nutritional status, functionality and quality of life of nursing home residents with malnutrition or at risk of malnutrition.

In 6 nursing homes in Nuremberg and Fuerth, Germany, a standardized screening was performed to identify all residents with malnutrition or at risk of malnutrition who might possibly benefit from nutritional intervention. All subjects with either a Mini Nutritional Assessment (MNA®) score below 24 points, BMI ≤22 kg/m², a low food intake according to the nurses' perception or weight loss of ≥5% in the last 3 or ≥10% in 6 months, respectively, were invited to participate and asked for informed consent.

Eligible residents with informed consent were randomly assigned to the intervention group (IG) which received two bottles of the ONS per day (250 mL, 600 kcal) in addition to usual nutritional care or to the control group (CG) which received usual nutritional care only.

Assessment of nutritional parameters included body weight, BMI, arm- and calf-circumference and Mini Nutritional Assessment (MNA). Cognitive status, depression, activities of daily living, handgrip strength, gait speed and quality of life were examined using standardized instruments and protocols. All measurements and tests were performed at baseline, after 12 and 24 weeks.

Compliance and tolerance were documented daily by nursing staff and regularly controlled by the study team.

Statistical analysis was performed following the intention-to-treat (ITT) approach including all residents originally assigned to either the IG or CG unless residents died during the study. A sample size calculation was performed with body weight as the primary outcome parameter (0.8 power to detect a significant difference p<0.05, two-sided) based on an estimated mean body weight 55 kg and a mean difference in body weight between the groups after 12 weeks of 1.5±2.2 kg comparable to results previously observed in this population. To detect a significant difference between IG and CG, 35 subjects were needed for each group. While sample size calculation aimed at ensuring adequate power to detect meaningful differences, the actual statistical analysis was of exploratory nature.