Insulin Resistance and N-Terminal Pro-B-Type Natriuretic Peptide Among Healthy Adults

Abstract

Importance: It is unclear to what extent insulin resistance is associated with N-terminal pro-B-type natriuretic peptide (NT-proBNP) in the general population after accounting for body composition.

Objective: To characterize the association of insulin resistance with NT-proBNP independently of measures of body composition in US adults.

Design, setting, and participants: In a cross-sectional design, data on participants aged 20 years or older were obtained from the 1999-2004 National Health and Nutrition Examination Survey with measures of NT-pro-BNP, body mass index (BMI), and dual energy x-ray absorptiometry (DEXA)-derived measures of body composition (fat and lean masses). Linear and logistic regression was used to characterize the associations of measures of body mass and composition (BMI, waist circumference, fat mass, and lean mass) with NT-proBNP, adjusting for cardiovascular risk factors. Linear regression was used to characterize the associations of homeostasis model assessment of insulin resistance [HOMA-IR] and NT-proBNP after adjusting for cardiovascular risk factors and body composition measures. The quantitative insulin sensitivity check index [QUICKI], triglyceride-glucose index [TyG index], insulin to glucose ratio [IGR], fasting insulin, and homeostasis model assessment of β-cell function (HOMA-β) were also examined. Data for this study were analyzed from August 10, 2022, to June 30, 2023.

Main outcomes and measures: Adjusted changes in NT-proBNP by insulin resistance levels.

Results: A total of 4038 adults without diabetes or cardiovascular disease were included (mean [SD] age, 44 years; 51.2% female; and 74.3% White). In sex-specific analyses, insulin resistance measures were inversely associated with NT-pro-BNP. After adjustment including cardiovascular risk factors, BMI, waist circumference, and DEXA-derived fat mass and lean mass, the percent change in NT-proBNP associated with an SD increase in HOMA-IR was -16.84% (95% CI, -21.23% to -12.21%) in women and -19.04% (95% CI, -24.14 to -13.59) in men. Similar associations were observed for other indices of insulin resistance, including QUICKI (women: 17.27; 95% CI, 10.92-23.99 vs men: 22.17; 95% CI, 15.27 to 29.48), TyG index women: -11.47; 95% CI, -16.12 to -6.57 vs men: -15.81; 95% CI, -20.40 to -10.95), IGR women: -15.15; 95% CI, -19.35 to -10.74 vs men: -16.61; 95% CI, -21.63 to -11.26), and fasting insulin (women: -16.32; 95% CI, -20.63 to -11.78 vs men: -18.22; 95% CI, -23.30 to -12.79), as well as HOMA-β (women: -10.71; 95% CI, -14.71 to -6.52 vs men: -11.72; 95% CI, -16.35 to -6.85).

Conclusions and relevance: In a national sample of US adults, insulin resistance was inversely associated with NT-proBNP, even after rigorously accounting for multiple measures of fat mass and lean mass. These results suggest that the mechanisms linking NT-proBNP to insulin resistance are partially independent of excess adiposity and may be associated with hyperinsulinemia.

Figure 1.. Associations of Measures of Insulin Resistance and Insulin Secretion With N-Terminal Pro-B-Type Natriuretic Peptide (NT-proBNP) Levels Among Women

Each insulin resistance/secretion or glycemic measure was modeled as sex-specific restricted cubic splines with knots at the 5th, 35th, 65th, and 95th percentiles. Model 1 adjusted for age, race and ethnicity, educational level, alcohol use, smoking, physical activity, hypercholesterolemia, hypertension, and estimated glomerular filtration rate (linear spline with knot at 60 mL/min/1.73 m²). Predictions were made with average values of all covariates. To convert glucose to millimoles per liter, multiply by 0.0555; HbA_1c to proportion of total hemoglobin, multiply by 0.01; insulin to picomoles per liter, multiply by 6.945. HbA_1c indicates hemoglobin A_1c; HOMA-β, homeostasis model assessment of β-cell function; HOMA-IR, homeostasis model assessment of insulin resistance; and QUICKI, quantitative insulin sensitivity check index.

Figure 2.. Associations of Measures of Insulin Resistance and Insulin Secretion With N-Terminal Pro-B-Type Natriuretic Peptide (NT-proBNP) Levels Among Men

Each insulin resistance/secretion or glycemic measure was modeled as sex-specific restricted cubic splines with knots at the 5th, 35th, 65th, and 95th percentiles. Base model adjusted for age, race and ethnicity, educational level, alcohol use, smoking, physical activity, hypercholesterolemia, hypertension, and estimated glomerular filtration rate (linear spline with knot at 60 mL/min/1.73 m²). Predictions were made with average values of all covariates. To convert glucose to millimoles per liter, multiply by 0.0555; HbA_1c to proportion of total hemoglobin, multiply by 0.01; insulin to picomoles per liter, multiply by 6.945. HbA_1c indicates hemoglobin A_1c; HOMA-β, homeostasis model assessment of β-cell function; HOMA-IR, homeostasis model assessment of insulin resistance; and QUICKI, quantitative insulin sensitivity check index.