Comparison of two nutritional assessment methods in gastroenterology patients

To investigate and compare efficacy and differences in the nutritional status evaluation of gastroenterology patients by application of two methods: subjective global assessment (SGA) and nutritional risk index (NRI).

The investigation was performed on 299 hospitalized patients, aged 18-84 years (average life span 55.57 +/- 12.84), with different gastrointestinal pathology, admitted to the Department of Gastroenterohepatology, Clinical and Hospital Center "Bezanijska Kosa" during a period of 180 d. All the patients, after being informed in detail about the study and signing a written consent, underwent nutritional status analysis, which included two different nutritional indices: SGA and NRI, anthropometric parameters, bioelectrical impedance analysis, and biochemical markers, within 24 h of admission.

In our sample of 299 hospitalized patients, global malnutrition prevalence upon admission varied from 45.7% as assessed by the SGA to 63.9% by NRI. Two applied methods required different parameters for an adequate approach: glucose level (5.68 +/- 1.06 mmol/L vs 4.83 +/- 1.14 mmol/L, F = 10.63, P = 0.001); body mass index (26.03 +/- 4.53 kg/m(2) vs 18.17 +/- 1.52 kg/m(2), F = 58.36, P < 0.001); total body water (42.62 +/- 7.98 kg vs 36.22 +/- 9.32 kg, F = 7.95, P = 0.005); basal metabolic rate (1625.14 +/- 304.91 kcal vs 1344.62 +/- 219.08 kcal, F = 9.06, P = 0.003) were very important for SGA, and lymphocyte count was relevant for NRI: 25.56% +/- 8.94% vs 21.77% +/- 10.08%, F = 11.55, P = 0.001. The number of malnourished patients rose with the length of hospital stay according to both nutritional indices. The discriminative function analysis (DFA) delineated the following parameters as important for prediction of nutritional status according to SGA assessment: concentration of albumins, level of proteins, SGA score and body weight. The DFA extracted MAMC, glucose level and NRI scores were variables of importance for the prediction of whether admitted patients would be classified as well or malnourished.

SGA showed higher sensitivity to predictor factors. Assessment of nutritional status requires a multidimensional approach, which includes different clinical indices and various nutritional parameters.