Abstract
Improving strategies for predicting the risk of developing heart failure (HF) and cardiovascular mortality in the older population is challenging. Previous studies use models that utilize troponin measurements using standard assays, but these lack sensitivity to detect troponin in the general population. This study analysed serum samples with the new high-sensitive troponin T (cTnT) assay in individuals recruited to the Cardiovascular Health Study (a multicentre prospective observational study of cardiovascular disease in community-dwelling ambulatory older adults).
cTnT concentrations were measured at baseline (n = 4221) and at 2–3 y intervals (n = 2918). Patients were followed up for a median of 11.8 y for outcomes of new-onset HF and cardiovascular death. Individuals were categorized based on baseline cTnT concentration: <3, 3–5.44, 5.45–8.16, 8.17–12.94 and >12.94 ng/L.
The incident rate (IR) for development of HF and cardiovascular death was associated with the baseline cTnT concentration (cTnT >12.94 ng/L: IR of 6.4 [per 100 y] and adjusted hazard ratio [aHR] of 2.48 for HF, IR of 4.8 and aHR of 2.91 for cardiovascular death; compared with cTnT <3 ng/L: IR of 1.6 and 1.1 for HF and cardiovascular death, respectively).
In individuals with detectable cTnT at baseline, an increase of >50% was associated with increased risk of HF and cardiovascular death (aHR: 1.61 and 1.65, respectively), while a decrease of >50% was associated with a decreased risk of HF and cardiovascular death (aHR: 0.71 and 0.73, respectively).
The addition of baseline cTnT to clinical risk factor models modestly improved classification and discrimination of patients with HF and cardiovascular death, and the additional inclusion of the change in cTnT further improved classification for HF and cardiovascular death.
The paper concludes that low concentrations of cTnT are associated with a gradient risk for new-onset HF and cardiovascular death in community-dwelling adults >65 y old, and that longitudinal changes in cTnT concentration are independently associated with a dynamic change in risk of HF and cardiovascular death.