Abstract 49: Changes in Circulating Natriuretic Peptide and Adiponectin Levels and All-cause Mortality in Older People: The Cardiovascular Health Study
Background: Despite its insulin-sensitizing and atheroprotective properties, higher plasma adiponectin (APN) has been linked to increased mortality in older adults. The basis for this adverse association is not well delineated, but natriuretic peptides (NPs) have been invoked as a potential explanation because they are known to stimulate APN secretion. One prospective study showed that adjustment for baseline NP levels attenuated APN’s relationship with mortality, but another demonstrated persistence despite such adjustment. The relationship between serial changes in NP and APN levels, however, and their association with fatal events, has not been previously examined.
Methods: We used linear regression to estimate the association of change (Δ) in N-terminal pro-B-type natriuretic peptide (NT-proBNP) with concurrent ΔAPN, and Cox regression to estimate the relative risk of death for ΔNT-proBNP and ΔAPN in a population-based study of older adults (Cardiovascular Health Study). NT-proBNP and APN were measured in stored plasma from the 1996 exam to complement previous measures of both analytes obtained at the 1992 exam. There were n=2669 participants with all measures; APN values were well harmonized across the 2 exams, but because NT-proBNP values >1000 pg/ml were not, we excluded such values (n=238) from the primary analysis. Both annualized change and average level across both years were considered in order to distinguish true change from improved precision from repeated measurement.
Results: Age 77±4 yrs, 62% women; ΔNT-proBNP 15±37 pg/mL/yr; ΔAPN 0.3±1.2 mg/L/yr. After full adjustment (age, sex, race, smoking, alcohol, Δweight, Δglucose, albumin, medications, prevalent CVD, health status, eGFR, and CRP), ΔNT-proBNP, but not mean NT-proBNP, was positively associated with ΔAPN (0.1 mg/L/yr [0.06, 0.16] APN increase per SD increment in ΔNT-proBNP). During follow-up through 2010, 1792 participants died. In fully adjusted models, both mean APN and ΔAPN were significantly associated with mortality (HR per SD increment 1.06 [1.00, 1.12] and 1.1 [1.05, 1.16], respectively), whereas mean NT-proBNP (1.25 [1.18, 1.33]), but not ΔNT-proBNP, was significantly related to this outcome. When both measures of APN & NT-proBNP were included in a fully adjusted model, mean APN ceased to be significant, but associations of ΔAPN and mean NT-proBNP with mortality were not materially changed. Sensitivity analysis including NT-proBNP >1000 pg/mL with winsorization at the upper 99th %ile of NT-proBNP showed similar findings.
Conclusion: These findings provide new insights into APN’s association with mortality in older adults, showing that (i) longitudinal change in NT-proBNP is associated with change in APN; and (ii) it is this longitudinal change in APN, independent of concurrent change or average level of NT-proBNP, but not the adipokine’s mean level, that is associated with increased mortality.
Author Disclosures: J.R. Kizer: None. P. Buzkova: None. A.M. Arnold: None. C. deFilippi: None. E.S. Strotmeyer: None. J.L. Sanders: None. R.C. Kaplan: None. M. Cushman: None. S.B. Kritchevsky: None. C.H. Hirsch: None. A.B. Newman: None.
- © 2015 by American Heart Association, Inc.