Scholarlink

Background <p>Diastolic dysfunction (DD) is a common feature in older adults, but its prognostic value is unclear.</p> Aim <p>This study aims to assess the ability of DD against NT-proBNP and frailty to predict all-cause mortality, cardiovascular mortality, and a first unplanned hospitalization in older adults.</p> Methods <p>Secondary analysis of the observational population-based BELFRAIL cohort of patients aged ≥ 80 years with cardiac echography at inclusion. Patients were included if LVEF ≥ 50% and without severe valvular disease. DD was defined if 50% of the criteria of the American Society of Echocardiography were fulfilled (average E/e’&gt;14, septal e’ velocity &lt; 7&#xa0;cm/s or lateral &lt; 10&#xa0;cm/s, tricuspid velocity &gt; 2.8&#xa0;m/s, left atrial volume index &gt; 34ml/m<sup>2</sup>), elevated NT-proBNP as a level ≥ 400ng/mL, and frailty as Fried Frailty Index ≥ 3. Multivariable Cox regression was used to adjust the hazard of mortality (all-cause and cardiovascular) and unplanned hospitalizations. A hierarchical decision tree was computed using classification and regression trees (CART).</p> Results <p>Of the 393 patients (mean age 85 years [SD 3,6], 257 (65%) women), 185 (47%) had DD, 76 (19%) an elevated NT-proBNP, and 50 (13%) were frail. During 5.1 ± 0.2 years, 143 (36%) patients died, of whom 55 (14%) from CV causes. Crude mortality was worse (log-rank <i>p</i> &lt; 0.05) for patients with DD (HR 1.48 [1.07–2.06]), elevated NT-proBNP (2.07 [HR 1.44–2.97]) or frailty (HR 3.02 [2.05–4.47]). After adjustment, DD predicted only CV mortality (HR 1.84 [1.03–3.31]). NT-proBNP predicted both all-cause (HR 1.52 [1.03–2.24]) and CV mortality (HR 2.16 [1.20–3.87]). Frailty predicted all-cause mortality (HR 2.10 [1.38–3.21]) and the first unplanned hospitalization (HR 1.62 [1.08 - 2.42]). Regarding CART, frailty was the root node for both predicting the risk of all-cause mortality and a first unplanned hospitalization. NT-proBNP was the root node in the CV mortality tree, followed by frailty.</p> Conclusion <p>Compared to NT-proBNP and frailty, DD offers limited added value in risk stratification for older people with preserved ejection fraction and no valvular disease. Frailty emerged as the strongest and most consistent predictor of mortality and hospitalization, and was central in the all-cause mortality and hospitalization decision-tree models.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Untangling the prognostic value of diastolic dysfunction, NT-proBNP, and frailty in older patients with preserved left ventricular ejection fraction without valvular disease

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月1日

查看全文
  • Christophe de Terwangne,
  • Bert Vaes,
  • Agnès Pasquet,
  • Benoit Boland,
  • Anne-Catherine Pouleur,
  • Jean-Marie Degryse

摘要

Background

Diastolic dysfunction (DD) is a common feature in older adults, but its prognostic value is unclear.

Aim

This study aims to assess the ability of DD against NT-proBNP and frailty to predict all-cause mortality, cardiovascular mortality, and a first unplanned hospitalization in older adults.

Methods

Secondary analysis of the observational population-based BELFRAIL cohort of patients aged ≥ 80 years with cardiac echography at inclusion. Patients were included if LVEF ≥ 50% and without severe valvular disease. DD was defined if 50% of the criteria of the American Society of Echocardiography were fulfilled (average E/e’>14, septal e’ velocity < 7 cm/s or lateral < 10 cm/s, tricuspid velocity > 2.8 m/s, left atrial volume index > 34ml/m2), elevated NT-proBNP as a level ≥ 400ng/mL, and frailty as Fried Frailty Index ≥ 3. Multivariable Cox regression was used to adjust the hazard of mortality (all-cause and cardiovascular) and unplanned hospitalizations. A hierarchical decision tree was computed using classification and regression trees (CART).

Results

Of the 393 patients (mean age 85 years [SD 3,6], 257 (65%) women), 185 (47%) had DD, 76 (19%) an elevated NT-proBNP, and 50 (13%) were frail. During 5.1 ± 0.2 years, 143 (36%) patients died, of whom 55 (14%) from CV causes. Crude mortality was worse (log-rank p < 0.05) for patients with DD (HR 1.48 [1.07–2.06]), elevated NT-proBNP (2.07 [HR 1.44–2.97]) or frailty (HR 3.02 [2.05–4.47]). After adjustment, DD predicted only CV mortality (HR 1.84 [1.03–3.31]). NT-proBNP predicted both all-cause (HR 1.52 [1.03–2.24]) and CV mortality (HR 2.16 [1.20–3.87]). Frailty predicted all-cause mortality (HR 2.10 [1.38–3.21]) and the first unplanned hospitalization (HR 1.62 [1.08 - 2.42]). Regarding CART, frailty was the root node for both predicting the risk of all-cause mortality and a first unplanned hospitalization. NT-proBNP was the root node in the CV mortality tree, followed by frailty.

Conclusion

Compared to NT-proBNP and frailty, DD offers limited added value in risk stratification for older people with preserved ejection fraction and no valvular disease. Frailty emerged as the strongest and most consistent predictor of mortality and hospitalization, and was central in the all-cause mortality and hospitalization decision-tree models.