Intersubject variability of blood analysis reference values: assessment of age and locality influence by means of a linear discriminant analysis model

• Main Author: Barreira, João C.M.
• Publication Date: 2013
• Other Authors: Costa, Roberto Germano, Pombo, Graça, Ferreira, Isabel C.F.R.
• Format: Article
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10198/9506
• topic: LDL-C; HDL-C; Triglycerides; Glycated hemoglobin; Biostatistics

Glycemic and lipidic profiles might be influenced by several factors. The fact that the age group might alter the most extensively evaluated lipidic and glycemic parameters is a more or less well accepted fact. To verify this empirical notion, 996 human subjects aged between 21 and 90 years from different localities were characterized according to age. To assess lipid profile, total cholesterol and cholesterol associated with lipoprotein fractions (c-LDL and c-HDL) and triglycerides were determined. Regarding glycemic profile, glucose and glycated hemoglobin were measured. The majority of the population had values of lipidic parameters fit into the reference values, presenting low or moderate risk for developing cardiovascular disease. Blood glucose was often far above the desirable, but this can be devalued due to the HbA1c values, which were overwhelmingly located in the normal range. The categorization of data in different age groups did not allow defining statistically significant differences. Despite the discriminant linear model was presented, the results indicate that age group did not act as a strong discriminant factor. Somehow unexpectedly, the most significant differences were found among the different localities, which tended to show a similarity according with their latitude. Furthermore, there were no significant correlations in the parameters associated with lipidic profile, but there was a direct correlation between glucose levels and HbA1c (glycemic parameters).