RESUMO
This case-control study was undertaken in 75 subjects categorized into 3 equal groups (A: diabetic subjects with macroalbuminuria, B: non-diabetic subjects with macroalbuminuria and C; control subjects). Serum erythropoietin (EPO) was estimated and analyzed in relation to hemoglobin levels in the three groups. The Pearson's coefficient (r) for hemoglobin and log natural EPO was significant for groups A (0.01), B (0.05) and C (0.01). Linear regression analysis of hemoglobin and log natural EPO showed significant differences between the study and control groups; however no significant difference could be demonstrated amongst the study groups. Hence, it was concluded that an inadequate EPO production occurs in renal failure, which accounts for the anaemia and diabetes does not confer an additional discrepancy in this mechanism over non-diabetic macroalbuminuria.