Cardiac ventricles release B-type na-triuretic peptide (BNP) in response to volume expansion and pressure overload; therefore, BNP concentration may be used as a biochemical marker of cardiac failure (1). BNP levels are high in diabetic subjects with left ventricular dysfunction (2). Because the diagnosis of cardiac failure may be an emergency (3), it is important to know whether the timing of BNP measurement (before or after a meal) affects the result. The effect of renal failure is also an important practical issue, since 25–40% of diabetic subjects have diabetic nephropathy. This study was designed to describe the changes elicited by chronic renal failure (CRF) and meal consumption on BNP levels in diabetic patients. Thirty diabetic patients were divided into two groups: 15 patients without CRF (group 1: glomerular filtration rate [GFR]) by Cockcroft’s formula 60 ml/min) and 15 patients with CRF (group 2: GFR 60 ml/min). No patient presented significant cardiac history based on clinical examination, electrocardiograph recording, and chest radiography. Four blood samples were collected before and 1, 2, and 3 h (T0 to T3) after a standardized meal (72 g carbohydrates, 21 g lipids, and 32 g proteins) to measure plasma BNP concentrations (IRMA Shionoria-BNP, Schering Cis Bio). Mean BNP levels (SD) for the two groups were compared using the Mann-Whitney U test. Values for repeated data before and after lunch were compared by ANOVA. Sex, age, BMI, type of diabetes, HbA1c, and blood pressure were similar in the two groups. GFR was lower in group 2 (group 1: 92 20, group 2: 36 15 ml/min; P 0.001). No significant difference was found between BNP concentrations before and after lunch in both groups (group 1: T 12.3 23.1, T1 11.7 21.9, T2 12.3 21.9, T3

12.5 23.2 pg/ml; group 2: T 34.9 37.9, T1 34.7 38.9, T2 35.9 36.7, T3 35.7 39.4 pg/ml). The BNP concentrations were higher in group 2 (P 0.05 for each time). When the two groups were analyzed together, BNP concentrations were negatively correlated with GFR (r 0.56, P 0.005). Natriuretic peptides play many physiological roles; therefore, determining whether feeding affects BNP concentrations is a practical problem. Plasma-atrial natriuretic peptide concentrations are increased by feeding in nondiabetic subjects (4). We show here that feeding does not affect plasma BNP levels in diabetic patients; blood samples can be drawn at any time for BNP measurement.

In nondiabetic patients, BNP distinguishes patients with a left ventricular hypertrophy and CRF (5). In our study, plasma BNP concentrations were higher in diabetic patients with CRF before and after feeding. Plasma BNP may therefore be used to monitor the cardiac function in diabetic subjects with CRF, but its reduced clearance participating in its increased concentrations needs to be taken into account.