Insulin resistance is associated with increased concentrations of NT-proBNP in rheumatoid arthritis: IL-6 as a potential mediator

Abstract

We examined the hypothesis that insulin resistance (IR) decreases circulating concentrations of N-terminal (NT)-probrain natriuretic peptide (BNP). Obesity, despite being a risk factor for heart failure (HF), is paradoxically associated with lower concentrations of BNP, a marker of myocardial stress. Low BNP in obesity is postulated to be due to IR; however, it has been difficult to define the role of IR independent of obesity. IR in rheumatoid arthritis (RA) is increased, independent of obesity, thus allowing potential mechanistic insights into the relationship between IR and BNP. We measured demographic factors, traditional cardiovascular risk factors, body mass index (BMI), markers of inflammation (interleukin-6 (IL-6), C-reactive protein (CRP), tumor necrosis factor α (TNFα)), NT-proBNP, and IR by the homeostatic model assessment (HOMA) in 140 patients with RA and 82 control subjects. Patients with heart failure and coronary artery disease were excluded. We used multiple linear regression models to examine the relationship between HOMA and NT-proBNP in RA and controls and in RA alone, the additional effect of inflammation. As previously reported, NT-proBNP concentrations were higher in RA (median 80.49 pg/mL, IQR (23.67-167.08 pg/mL)) than controls (17.84 pg/mL (3.28-36.28 pg/mL)) (P < 0.001), and the prevalence of IR, defined by HOMA > 2.114, was higher among RA than controls (53 % vs. 15%, P > 0.001). HOMA was positively correlated with NT-proBNP (rho = 0.226, P = 0.007) in RA, but not in controls (rho = -0.154, P = 0.168). In a multivariable model adjusted for age, race, and sex, we found that increasing HOMA was statistically associated with increasing NT-proBNP concentrations in RA (P = 0.001), but not controls (P = 0.543) (P for interaction = 0.036). In RA subjects, when IL-6 was further included in the model, IL-6 (P = 0.0014), but not HOMA (P = 0.43), remained significantly associated with NT-proBNP, suggesting that IL-6 may be mechanistically involved in the relationship between IR and NT-proBNP in RA. We conclude that in patients with RA, insulin resistance is associated with higher, rather than the expected lower, concentrations of NT-proBNP and that this may be related to increased IL-6.

Figure 1. Non-linear smooth regression is used to illustrate the relationships between HOMA-IR and NT-pro-BNP in patients with RA patients and controls

Analysis using a multiple linear regression model that adjusted for age, race, and sex, showed that HOMA-IR was positively associated with NT-pro-BNP in patients with RA (P=0.0014), but not in control subjects (P=0.543), (P for interaction=0.036).

Figure 2. Non-linear smooth regression is used to illustrate the relationships between BMI and NT-pro-BNP in patients with RA and controls

Analysis using a multivariable regression model that adjusted for age, race, and sex, showed that there was no statistically significant association between BMI and NT-pro-BNP in patients with RA or controls (both P values >0.2) (P for interaction=0.19).

Figure 3. Multivariable regression analyses examining the effect of inflammatory mediators on the relationship between HOMA and NT-pro-BNP in patients with RA

The relationship between HOMA and NT-pro-BNP, adjusted for age, race and sex in RA patients (P=0.0012) remained statistically significant after additional adjustment TNFα (P=0.015) and CRP (P=0.008). However, adjustment for IL-6 rendered the association between HOMA and NT-pro-BNP non-significant (P=0.43).