<p>Community-acquired pneumonia (CAP) remains a leading cause of death in older adults, largely because immunosenescence and inflamm-aging obscure early clinical deterioration. We explored the short-term temporal interplay between immune and inflammatory biomarkers and assessed how age, malnutrition, and diabetes modify these dynamics and their prognostic value. A retrospective cohort study was conducted on 418 elderly CAP patients (≥ 65&#xa0;years). Lymphocyte counts, C-reactive protein (CRP), and procalcitonin (PCT) were recorded at admission (T0) and 24&#xa0;h intervals (T1, T2). Biomarker trends were analyzed using auto-regressive integrated moving average (ARIMA) models, while cross-correlation functions (CCF) quantified lagged relationships, and time-varying Cox models and time-dependent receiver-operating-characteristic (td-ROC) curves evaluated 28&#xa0;day mortality. Overall mortality was 18.4% (77/418). Lymphocyte counts were consistently lower in non-survivors (p &lt; 0.001). Falls in lymphocytes preceded rises in CRP and PCT by 24–48&#xa0;h. ARIMA projections mirrored observed trends and identified sustained lymphopenia and escalating PCT in non-survivors. In time-varying Cox analysis, lymphocyte count was the strongest protective factor (HR per 1 × 10<sup>9</sup>L<sup>−1</sup> increase = 0.45; 95% CI 0.35–0.58; p &lt; 0.001), with amplified impact in malnourished, diabetic and ≥ 85&#xa0;years subgroups. td-ROC showed PCT possessed the highest discrimination, while lymphocyte AUC rose over time and approached PCT. Malnutrition and diabetes attenuated CCF strengths, and patients ≥ 85&#xa0;yr displayed blunted immune-inflammatory coupling. Early lymphocyte collapse foretells subsequent inflammatory surges and adverse outcomes in elderly CAP. Malnutrition, diabetes, and advanced age disrupt this immune-inflammatory synchrony, diminishing prognostic precision. Routine 72&#xa0;h serial monitoring, coupled with dynamic modelling, may enable pre-emptive nutritional and metabolic interventions and refine risk-stratified management.</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

Time series modeling shows early lymphocyte decline predicts inflammatory rise and mortality in older adults with community acquired pneumonia

  • Jingxian Liao,
  • Chunhui Xie,
  • Lei Miao,
  • Yunfeng Li,
  • Chen Gong,
  • Qing Xiao

摘要

Community-acquired pneumonia (CAP) remains a leading cause of death in older adults, largely because immunosenescence and inflamm-aging obscure early clinical deterioration. We explored the short-term temporal interplay between immune and inflammatory biomarkers and assessed how age, malnutrition, and diabetes modify these dynamics and their prognostic value. A retrospective cohort study was conducted on 418 elderly CAP patients (≥ 65 years). Lymphocyte counts, C-reactive protein (CRP), and procalcitonin (PCT) were recorded at admission (T0) and 24 h intervals (T1, T2). Biomarker trends were analyzed using auto-regressive integrated moving average (ARIMA) models, while cross-correlation functions (CCF) quantified lagged relationships, and time-varying Cox models and time-dependent receiver-operating-characteristic (td-ROC) curves evaluated 28 day mortality. Overall mortality was 18.4% (77/418). Lymphocyte counts were consistently lower in non-survivors (p < 0.001). Falls in lymphocytes preceded rises in CRP and PCT by 24–48 h. ARIMA projections mirrored observed trends and identified sustained lymphopenia and escalating PCT in non-survivors. In time-varying Cox analysis, lymphocyte count was the strongest protective factor (HR per 1 × 109L−1 increase = 0.45; 95% CI 0.35–0.58; p < 0.001), with amplified impact in malnourished, diabetic and ≥ 85 years subgroups. td-ROC showed PCT possessed the highest discrimination, while lymphocyte AUC rose over time and approached PCT. Malnutrition and diabetes attenuated CCF strengths, and patients ≥ 85 yr displayed blunted immune-inflammatory coupling. Early lymphocyte collapse foretells subsequent inflammatory surges and adverse outcomes in elderly CAP. Malnutrition, diabetes, and advanced age disrupt this immune-inflammatory synchrony, diminishing prognostic precision. Routine 72 h serial monitoring, coupled with dynamic modelling, may enable pre-emptive nutritional and metabolic interventions and refine risk-stratified management.