Autor:innen:
D. Dallmeier (Ulm, DE)
U. Braisch (Ulm, DE)
W. Koenig (München, DE)
D. Rothenbacher (Ulm, DE)
M. Denkinger (Ulm, DE)
Background: N-Terminal pro Brain Natriuretic Peptide (NT-proBNP) is released from the ventricular myocardium in the settings of volume and/or pressure overload, and is recognized as a good marker for myocardial hemodynamic stress. We evaluated the association of established predictors of NT-proBNP across its quantiles in older people.
Methods: NT-proBNP levels were measured in 1506 community-dwelling adults ≥65 years, participants at the longitudinal cohort study Activity and Function in the Elderly in Ulm study – ActiFE Ulm. Quantile regression evaluated the association of age, sex, body mass index (BMI), smoking, cystatin C based glomerular filtration rate (GFR), hemoglobin, testosterone, C-reactive protein (CRP), history of hypertension, diabetes, myocardial infarction, heart failure, and stroke with NT-proBNP levels measured at baseline at the 0.10, 0.25, 0.50, 0.75 and 0.90 quantiles.
Results: In our study population (n=1456, mean age 75.5 years, 56.9% male) NT-proBNP was not normally distributed with a median level of 153.0 ng/L [interquartile range Q1 81.9, Q3 318.0]. Simple quantile regressions showed a significant association for age, sex, GFR, hemoglobin, testosterone, CRP, history of hypertension, myocardial infarction, heart failure and stroke. In the multivariable analysis we observed a steady significant increment of the ß-estimates across the 0.10, 0.25, 0.50, 0.75, 0.90 quantiles for following variables: age (ß 2.2, 3.5, 5.7, 11.8, 17.9), CRP (ß 1.3, 3.4, 6.9, 8.2, 14.2), hypertension (ß 3.0, 18.7, 34.7, 63.7, 80.7), myocardial infarction (ß 49.2, 96.4, 213.5, 290.7, 684.8), and heart failure (ß 32.8, 49.7, 106.9, 418.5, 704.7). In addition increments of BMI (ß -1.3, -3.0, -5.1, -5.7, -12.9), GFR (ß -0.5, -1.0, -1.8, -3.9, -9.8) and hemoglobin (ß -7.7, -8.9, -19.4, -32.4, -30.1) were significantly associated with lower levels of NT-proBNP across the quantiles. No statistically significant association could be detected for sex, testosterone, smoking, history of diabetes and of stroke.
Conclusion: Our data showed not only that age, BMI, GFR, hemoglobin, CRP, history of hypertension, myocardial infarction and stroke are associated with the baseline levels of NT-proBNP after further adjustment for sex, smoking, testosterone, and history of diabetes, but also that their level of contribution varies considerably across the 0.10, 0.25, 0.50, 0.75 and 0.90 quantiles, with the higher ß-estimates observed by history of myocardial infarction, heart failure, and hypertension. These results emphasize the need to deepen our understanding about the pathophysiology of NT-proBNP, so that the assumption of a linear relationship between these established predictors and NT-proBNP in older people should be treated with the utmost caution. Next steps should analyze the effects of these predictors across quantiles of NT-proBNP changes over time, allowing the identification of those factors with the most beneficial or rather detrimental effect.