The present study demonstrated that plasma BNP level examined during medical checkups of the general Japanese population is a strong independent predictor of MACE. The addition of BNP to the medical checkup parameters may improve the quality cardiovascular disease prediction and, thereby, may play a role in cardiovascular disease prevention.
BNP is synthesized and secreted from ventricular cells mainly in response to an increase in ventricular volume or pressure [9–13]. There is increasing evidence that circulating levels of plasma BNP is a reliable marker of left ventricular function and the severity and prognosis of heart failure [11–13]. However, careful interpretation is necessary because numerous factors, such as sex, age, blood pressure, kidney function, and obesity, seriously affect BNP levels [11–15]. Furthermore, the clinical significance of relatively low levels of peptide seems to be obscure. Although the normal reference value for BNP level is 18.4 pg/ml or less, values above the normal reference value do not necessarily indicate heart failure. The Japanese Heart Failure Society set a cut-off point of 40 pg/ml for plasma BNP concentration to identify patients with mild heart failure and of 100 pg/mg to identify patients with heart failure requiring medication [16]. Thus, a BNP predictive value that is less than 40 pg/ml been considered low for identifying heart failure. The median value and interquartile range of BNP in the present study was 10.7 and 6.4–17.8 pg/ml, respectively, which clearly demonstrates the clinical significance of relatively low levels of BNP in the general population. The present results are consistent with that of a previous study demonstrating that relatively low levels of BNP were associated with death and various cardiovascular diseases [17]. It is noteworthy that BNP measured during medical checkup predicts MACE independent of classic risk factors, such as hypertension, diabetes mellitus, dyslipidemia, and smoking. This present finding implies that BNP and classic risk factors play a complementary role in the prediction and, thereafter, prevention of cardiovascular events. The results of the sub-analysis conducted on participants without medication that may affect the cardiovascular system further support the concept that BNP is a strong independent predictor of MACE. However, careful interpretation is necessary because more than a half of participants with hypertension, diabetes mellitus, or dyslipidemia were excluded from the sub-analysis and blood pressure, glucose tolerance, and lipid profile were nearly normal in most of the participants included in the sub-analysis.
The mechanism underlying the association between BNP and future occurrence of MACE was not elucidated in the present study. The prediction of heart failure using BNP levels is easy to understand when considering the mechanism of BNP secretion [9–13]. Latent left ventricular dysfunction induced by subclinical or concealed myocardial ischemia may have increased BNP levels, resulting in its prediction of clinical myocardial ischemia, such as myocardial infarction, unstable angina, or angina requiring hospitalization for percutaneous coronary intervention. In line with the speculation, BNP is predictive of silent myocardial ischemia in patients with non-obstructive hypertrophic cardiomyopathy [18], and BNP predicts recurrence of angina pectoris [19]. Stroke was the most frequent MACE in the present study. Recent studies have suggested that salt intake causes mild BNP elevation (below 30 pg/ml) in the general population [15, 20] and that excessive salt intake is an independent risk factor for cardiovascular events [21]. Thus, the mild increase of BNP observed in the present study may at least partially reflect excessive salt intake. BNP is also elevated in patients with hypertension, and increasing blood pressure stimulates BNP secretion [22], suggesting that an increase in BNP may have partially occurred following mild blood pressure elevation although adjustment for blood pressure in Cox regression analysis did not eliminate the significant correlation between BNP and MACE.
The present study confirmed the importance of the identification of individuals with metabolic syndrome as well as other classic risk factors. Although waist circumference, high-density lipoprotein cholesterol, or triglyceride alone was not an independent risk factor for MACE in the present observational study, a cluster of mild metabolic disorders independently predict future cardiovascular events. Low-density lipoprotein cholesterol or current smoking habit did not show significant correlation with MACE. Relatively low levels of low-density lipoprotein cholesterol in the present study participants and the presence of ex-smokers may have affected the results, although detailed mechanisms are uncertain.
The interpretation of the present results is limited by the following points. The participants were participants in our annual medical checkup program, and selection bias is possible. The outcome was confirmed by telephone or letter in some participants who did not visit our hospital after baseline examination, leading to a likelihood of an inaccurate diagnosis of the outcome in those participants. Although the large number of participants included in the present study may overcome some of these limitations, a study with more participants and longer observation period is necessary to arrive at a definite conclusion.