A Novel Hormone Deficiency?  Investigation of “Natriuretic Peptide Deficiency” in a Large, De-Identified Clinical Database

Presentation Number: SUN 505
Date of Presentation: April 2nd, 2017

Katherine Neubecker Bachmann*1, Deepak K Gupta2, Quinn S Wells2, Eric Farber-Eger2, Evan Brittain2 and Thomas J Wang2
1Vanderbilt University Medical Center, Nashville, TN, 2Vanderbilt University Medical Center


Background: The natriuretic peptide (NP) hormonal system plays a crucial role in the regulation of blood pressure and volume status. In response to cardiac wall stress, cardiomyocytes release NPs, which cause natriuresis, vasodilation, and inhibition of the renin-angiotensin-aldosterone system. Circulating NP levels are particularly high in the setting of cardiac dysfunction or overt heart failure (HF), in which B-type NP (BNP) levels can far exceed 400 pg/ml1. Although NPs are hormones, it is unknown whether states of “NP deficiency” can exist, analogous to other hormonal deficiencies such as adrenal insufficiency or hypothyroidism. We postulated that BNP levels <100 pg/ml in the setting of HF are inappropriately low and suggest NP deficiency. Thus, we examined a large, de-identified clinical database to determine the prevalence of low BNP levels among patients with overt HF or cardiac dysfunction.

Methods: We used the Vanderbilt Synthetic Derivative, a database of more than 2 million de-identified electronic medical records of patients treated at Vanderbilt University Medical Center. Among adults (≥18 years) who had BNP measurements, we identified 3 groups of patients with overt HF or cardiac dysfunction: 1) overt clinical HF requiring hospitalization [n= 9234]; 2) abnormal cardiac function or structure by echocardiography (with ejection fraction ≤35% or left ventricular (LV) hypertrophy) [n= 7323]; 3) abnormal hemodynamics by cardiac catheterization (with wedge, LV end diastolic, or right atrial pressure ≥ 20 mmHg; or cardiac index < 2 L/min/m2) [n= 523].

Results: The prevalence of BNP concentrations <100 pg/ml was 11% in patients with overt HF, 25% in patients with abnormal echocardiograms, and 29% in patients with abnormal hemodynamics. Very low BNP levels (<50 pg/ml) were observed in 5%, 14%, and 16% of patients with overt HF, abnormal echocardiograms, or abnormal hemodynamics, respectively. Patients with BNP<100 pg/ml had a much higher prevalence of obesity (69%, 62%, and 48% in those with overt HF, abnormal echocardiograms, and abnormal hemodynamics, respectively) than those with appropriately elevated BNP (37%, 33%, and 32%, respectively; all p<0.001). The low BNP subset also included more African-Americans compared to those with higher BNP (overt HF group: 21% vs 16%, p<0.0001; abnormal echocardiogram group: 22% vs 16%, p<0.001). In multivariable regression analyses adjusted for age, sex, and race, low BNP remained significantly associated with higher BMI.

Conclusions: A substantial minority of patients with overt HF or cardiac dysfunction have BNP concentrations below 100 pg/ml. Because BNP levels in this range are inappropriately low in the setting of volume overload or elevated cardiac wall stress, our findings suggest the possibility of “NP deficiency” in some individuals. Further investigations are warranted to establish the clinical consequences of NP deficiency.


Nothing to Disclose: KNB, DKG, QSW, EF, EB, TJW