Evaluation of nutritional status in patients with end-stage renal disease in hemodialysis using principal component analysis

Abstract

The evaluation of nutritional status in patients with end-stage renal disease in hemodialysis is composed
of a large number of measurements that complicate their execution. Therefore, the objective of this study
is to reduce the number of variables through the principal component analysis (PCA). For this, a PCA was
performed with 10 variables of the nutritional diagnosis in patients with hemodialysis: Energy Intake, Protein
Intake, IBM,% UBW,% SBW,% MUAC, cAMA,% TCF, HGS and TLC as well as the age of the patients. The
results show that PCA matrix with orthogonal rotation Varimex yielded four main components of the evaluation
of the nutritional status of renal disease in patients with end-stage hemodialysis, whose value was greater
than 0 and explains the 79.91% of the total variance. The first factor was called body composition status, which
is composed of cAMA (r = 0.9138), IBM (r = 0.8755),% MUAC (r = 0.8681) and% SBW (r = 0.6238). In the
second factor called nutritional risk, a correlation was observed with energy intake (r = -0.8934), protein intake
(r = -0.8752) and %TCF (0.5040). The third component called functional status risk is composed of age (r =
0.9022) and HGS (r = 0.8508). The fourth factor, called body composition stability, was correlated with %UBW
(r = 0.7456) and %TCF (r = 0.5825). The results of this study will allow reducing the number of variables for the
preparation of a nutritional diagnosis in hemodialysis patients. From many to one of the four main components:
1) body composition status, 2) nutritional risk, 3) functional status risk or 4) body composition stability.