In adults with severe obesity, NT-proBNP levels were independently associated with E/E' (β=0.913, p=0.011), HOMA-IR (β=-0.327, p=0.008), and creatinine (β=-1.251, p=0.004).
Cross-Sectional (n=97)
In adults with severe obesity, NT-proBNP levels are independently associated with diastolic function (E/E'), metabolic status (insulin resistance), and renal function, highlighting the need for context-specific interpretation.
Effect estimate: β = 0.913 for ln(E/E')
p-value: p=0.011
Abstract Introduction Cardiac function in individuals with severe obesity is influenced by interrelated structural, hemodynamic, and metabolic factors. NT-proBNP, a biomarker released in response to myocardial wall stress, is widely used for diagnosis, treatment monitoring, and prognostication in heart failure. However, its interpretation in obese populations is challenging, as NT-proBNP levels are often lower than expected despite cardiac dysfunction. The explanation of this paradox remains incompletely understood. Purpose To identify independent echocardiographic and metabolic determinants of NT-proBNP concentrations in adults with severe obesity referred for bariatric surgery. Methods Baseline characteristics were summarized, including anthropometric (BMI, BSA), hemodynamic (SBP and DBP, HR), echocardiographic (cardiac output, stroke volume, left ventricular mass index LVMI, anteroposterior left atrial diameter, mitral inflow velocities E and A, tissue Doppler-derived E′ velocity, and E/E′ ratio), metabolic (glucose, insulin, homeostatic model assessment of insulin resistance HOMA-IR, CRP, and renal (creatinine, cystatin C, eGFR) parameters. Skewed variables (NT-proBNP, CRP, HOMA-IR, creatinine, E/E′) were log-transformed to improve model fit. Univariable linear regressions were performed with log-transformed NT-proBNP (lnNTproBNP) as the dependent variable. Variables with p 0.10 or clinical relevance were entered into a multivariable model using stepwise selection (entry p 0.05, removal p 0.10). Model performance was assessed using R², adjusted R², Akaike and Bayesian information criteria (AIC, BIC), and variance inflation factors (VIFs). Results Among 97 patients (mean age 45.0 ± 9.5 years, BMI 43.1 ± 9.0 kg/m²), the median NT-proBNP was 62.2 pg/mL. In univariable analyses, higher lnNTproBNP was associated with age (β = 0.030, p = 0.001), ln(E/E′) (β = 1.28, p 0.001), LA/BSA (β = 0.193, p = 0.001), and A-wave velocity (β = 2.03, p = 0.001), while heart rate (β = –0.022, p = 0.007), insulin (β = –0.018, p = 0.010), lnHOMA-IR (β = –0.297, p = 0.025), and ln(creatinine) (β = –0.847, p = 0.077) were inversely associated. In the final multivariable model, age (β = 0.025, p = 0.017) and ln(E/E′) (β = 0.913, p = 0.011) remained positively associated, while lnHOMA-IR (β = –0.327, p = 0.008) and ln(creatinine) (β = –1.251, p = 0.004) remained inversely associated. The model explained 34.4% of the variance (adjusted R² = 0.30). Conclusions In adults with severe obesity, NT-proBNP levels were independently associated with diastolic function, metabolic status, and renal function. E/E′ was the strongest predictor, supporting its clinical utility in assessing left ventricular filling pressure. The inverse associations with insulin resistance and renal impairment highlight the need for context-specific interpretation of NT-proBNP in this population.
Szmigielski et al. (Thu,) conducted a cross-sectional in Severe obesity (n=97). Echocardiographic and metabolic parameters (E/E', HOMA-IR, creatinine) was evaluated on Log-transformed NT-proBNP concentration (lnNTproBNP) (β = 0.913 for ln(E/E'), p=0.011). In adults with severe obesity, NT-proBNP levels were independently associated with E/E' (β=0.913, p=0.011), HOMA-IR (β=-0.327, p=0.008), and creatinine (β=-1.251, p=0.004).