Effect of various weight loss interventions on serum NT-proBNP concentration in severe obese subjects without clinical manifest heart failure

Abstract

Obesity is associated with a "natriuretic handicap" indicated by reduced N-terminal fragment of proBNP (NT-proBNP) concentration. While gastric bypass surgery improves the natriuretic handicap, it is presently unclear if sleeve gastrectomy exhibits similar effects. We examined NT-proBNP serum concentration in n = 72 obese participants without heart failure before and 6 months after sleeve gastrectomy (n = 28), gastric bypass surgery (n = 19), and 3-month 800 kcal/day very-low calorie diet (n = 25). A significant weight loss was observed in all intervention groups. Within 6 months, NT-proBNP concentration tended to increase by a median of 44.3 pg/mL in the sleeve gastrectomy group (p = 0.07), while it remained unchanged in the other groups (all p ≥ 0.50). To gain insights into potential effectors, we additionally analyzed NT-proBNP serum concentration in n = 387 individuals with different metabolic phenotypes. Here, higher NT-proBNP levels were associated with lower nutritional fat and protein but not with carbohydrate intake. Of interest, NT-proBNP serum concentrations were inversely correlated with fasting glucose concentration in euglycemic individuals but not in individuals with prediabetes or type 2 diabetes. In conclusion, sleeve gastrectomy tended to increase NT-proBNP levels in obese individuals and might improve the obesity-associated "natriuretic handicap". Thereby, nutritional fat and protein intake and the individual glucose homeostasis might be metabolic determinants of NT-proBNP serum concentration.

Figure 1

Comparison of weight loss and concomitant changes in fasting blood glucose and NT-proBNP concentration after sleeve gastrectomy, gastric bypass surgery, and a 26-week weight loss program. Left column denotes pre- and post (A) weight, (D) fasting blood glucose, and (G) NT-proBNP concentration after 26 weeks of sleeve gastrectomy follow-up. Middle column denotes pre- and post (B) weight, (E) fasting blood glucose, and (H) NT-proBNP concentration after 26 weeks of gastric bypass surgery follow-up. Right column denotes pre- and post (C) weight, (F) fasting blood glucose, and (I) NT-proBNP concentration after 13 weeks of VLCD + 13 weeks of weight maintenance. All three intervention groups were not statistically different in terms of baseline NT-proBNP concentration (p = 0.72). In panels (A–F), bars denote mean weight and error bars denote 95% CI of the mean. Statistical significance of changes in weight and fasting blood glucose from pre- to post-intervention was determined by Student’s paired t-test. Only complete and valid fasting blood glucose values were included in the analysis (sleeve gastrectomy: n = 14, gastric bypass n = 9, VLCD: n = 25). In panels G–I, bars denote median NT-proBNP concentration and error bars denote 95% CI of the median. Y axes are formatted in log₁₀ to account for skewed distribution of NT-proBNP values. Only NT-proBNP values above the ELISA detection limit were included in the analysis (sleeve gastrectomy: n = 14, gastric bypass n = 7, VLCD: n = 9). Statistical significance of changes in NT-proBNP concentration from pre- to post-intervention was determined by Wilcoxon signed rank test. CI; confidence interval; NT-proBNP, N-terminal pro brain natriuretic peptide; VLCD, very low-calorie diet.

Figure 2

Definition of NT-proBNP groups based on NT-proBNP plasma concentration in cross-sectional analysis. Due to the skewed distribution of NT-proBNP values in the cross-sectional FoCus cohort, we created three groups representing (1) NT-proBNP concentration below detection limit of 25.4 pg/mL (n = 220, left bar) as well as (2) low and (3) high NT-proBNP concentration. The latter two groups were created based on the median (158.9 pg/mL) of all detectable values. Individuals with an NT-proBNP concentration above 25.4 pg/mL but below 158.9 pg/mL were assigned to the low NT-proBNP group (n = 83, middle bar), while individuals with an NT-proBNP concentration equal or above 158.9 pg/mL were assigned to the high NT-proBNP group (n = 84, right bar). Bars denote median NT-proBNP concentration and error bars denote 95% CI of the median. Y axis is formatted in log₁₀ to account for skewed distribution of NT-proBNP values. CI; confidence interval; NT-proBNP, N-terminal pro brain natriuretic peptide.

Figure 3

Associations between NT-proBNP concentration and (A) body weight, (B) BMI, (C) age, and (D) systolic blood pressure. Bars denote mean and error bars denote 95% CI of the mean. P trend analyses were performed using a linear regression model with NT-proBNP group as ordinal covariate and the variable of interest as dependent variable. Definition of NT-proBNP groups is shown in Fig. 2. BMI, body mass index; CI; confidence interval; NT-proBNP, N-terminal pro brain natriuretic peptide.

Figure 4

Associations between NT-proBNP concentration and total energy and macronutrient intake. Association between NT-proBNP concentration and (A) energy intake, (B) carbohydrate intake, (C) fat intake, and (D) protein intake. Information about the nutritional intake was assessed by a self-reported 12-month nutritional, retrospective (food frequency) questionnaire. Bars denote mean and error bars denote 95% CI of the mean. P trend analyses were performed using a linear regression model with NT-proBNP group as ordinal covariate and the variable of interest as dependent variable. Definition of NT-proBNP groups is shown in Fig. 2. CI; confidence interval; NT-proBNP, N-terminal pro brain natriuretic peptide.

Figure 5

Subgroup analysis: associations between NT-proBNP concentration and fasting plasma glucose in (A) euglycemic, (B) prediabetic, and (C) diabetic individuals. Euglycemia was defined as fasting plasma glucose < 100 mg/dL, prediabetes was defined as fasting plasma glucose between 100 and 126 mg/dL, and diabetes was defined as fasting plasma glucose > 126 mg/dL. Bars denote mean and error bars denote 95% CI of the mean. P trend analyses were performed using a linear regression model with NT-proBNP group as ordinal covariate and the variable of interest as dependent variable. Statistical significance of between-group differences was calculated by unpaired t test. Definition of NT-proBNP groups is shown in Fig. 2. CI; confidence interval; NT-proBNP, N-terminal pro brain natriuretic peptide.