Abstract 5010: Determining the Plasticity of Cardiovascular Risk in the Elderly: A role for serial NT-proBNP measurement
In the elderly, with decades of exposure to cardiovascular (CV) risk factors, CV prognosis is often considered static. We hypothesized that CV risk is dynamic reflected by changes in NTproBNP over time. We measured NTproBNP at baseline and after 2–3 years in the Cardiovascular Health Study. Long-term risk of CV death according to change in NTproBNP was estimated with the Kaplan-Meier method. We used Cox models to test if effect of change in NTproBNP was independent of demographic and CV risk factors, baseline NTproBNP, renal function, coronary disease, and CV medications. Participants were categorized at baseline as < (low) or >=190pg/mL (high) levels, based on an observed increase in risk above this level. A significant change in NTproBNP category was defined as a change of >25% to a level above or below this cut-point, based on the reported biological variability of NTproBNP. Change in NTproBNP was also evaluated as a continuous measure. Serial NTproBNP levels were measured in 2,975 (86%) of 3,469 participants (age 75±5 years) without heart failure and who had a follow-up visit. CV death was different between those with low levels that remained low (n=1,774) vs. those with low levels that became high (n=468) (1.1 vs. 2.7 per 100 person-yrs, p<.001) and those with high levels that remained high (n=621) vs. those with high levels that became low (n=112) (4.2 vs. 1.6 per 100 person-yrs, p<.001) (figure⇓). As a continuous measure, change in NTproBNP was linearly associated with CV mortality risk after adjustment (per Ln-fold increment: RR=1.47, p<.001). Dynamic changes in NTproBNP levels reflect dramatic change in CV prognosis in the elderly.