a w y r b G h t a p n N r N w a t r N S t t o b G G s s v s a l The present research letter aims to compare a recently alidated index of nutritional risk, the Geriatric Nutritional isk Index (GNRI), with the Nutritional Risk Index (NRI) s they relate to biochemical and anthropometric variables sually investigated during the assessment of nutritional tatus. Moreover, we will suggest a possible use of the NRI in screening and grading nutritional status. The wide spectrum of tools available for nutritional asessment is growing larger. The GNRI, a new index for valuating at-risk elderly medical patients, was recently resented by Bouillanne et al. [1]. The GNRI is an adaptaion of the NRI, which was first described by Buzby et al. 2] to score nutritional risk in surgical patients, and is a imple and accurate, validated tool for predicting elderly atients at risk of morbidity and mortality. Because of the requent difficulty of obtaining usual body weight in elderly atients, Bouillanne et al. [1] hypothesized that this value in he NRI formula could be replaced by ideal body weight alculated according to the Lorentz formula, which in turn vercomes the unavailability of real height by using knee eight (NRI [1.519 albumin, g/L] [41.7 present/ sual body weight]; GNRI [1.489 albumin, g/L] 41.7 present/ideal body weight]). Both the GNRI and he NRI formulas are structured to provide greater weight to lbumin, which is a stronger predictor of mortality in the eneral population [3] than is body weight. However, the RI, which takes into account usual body weight, seems to e related more to a history of recent weight loss, an indiator that also has been associated with increased mortality 4]. In an extension of this observation, we present inforation about the relations between these two indexes of utritional risk and other biochemical and anthropometric