Integrated machine learning and population attributable fraction analysis of systemic inflammatory indices for mortality risk prediction in diabetes and prediabetes

Chronic systemic inflammation is a key contributor to cardiometabolic complications in diabetes mellitus (DM) and prediabetes (PreDM). Composite inflammatory indices—including neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), systemic inflammation response index (SIRI), systemic immune-inflammation index (SII), platelet-to-hemoglobin ratio (PHR), and aggregate inflammation systemic index (AISI)—have shown prognostic value for mortality. However, their integrated assessment using machine learning and quantification at the population level remain limited. In this retrospective cohort study, 11,304 adults with DM or PreDM from the National Health and Nutrition Examination Survey (NHANES, 2005–2018) were analyzed. The primary outcomes were all-cause and cardiovascular mortality. Associations between inflammatory indices and mortality were evaluated using Cox proportional hazards models. Predictive performance was assessed via Extreme Gradient Boosting (XGBoost), and population attributable fractions (PAFs) estimated the mortality burden related to systemic inflammation. NLR, MLR, SIRI, SII, and AISI were independently associated with all-cause and cardiovascular mortality. MLR showed the strongest association (HR: 2.948 and 3.717 for all-cause and CVD mortality, respectively). XGBoost identified SIRI, SII, AISI, MLR, and NLR as key predictors, with SIRI ranked highest for cardiovascular mortality. Inclusion of inflammatory indices improved model discrimination and calibration. PAF analysis suggested that 10–20% of mortality reduction could be attributed to improved inflammatory profiles. Systemic inflammatory indices are independent predictors of mortality in individuals with DM or PreDM. Their integration into machine learning models enhances risk prediction and may inform population-level strategies for cardiometabolic risk stratification. What is currently known about this topic?Chronic inflammation contributes to the progression of DM or PreDM.Traditional inflammatory markers are costly and less accessible for widespread screenings.Systemic inflammatory indices provide an easily accessible alternative for diagnosis and prognosis. Chronic inflammation contributes to the progression of DM or PreDM. Traditional inflammatory markers are costly and less accessible for widespread screenings. Systemic inflammatory indices provide an easily accessible alternative for diagnosis and prognosis. What is the key research question?How does machine learning improve mortality risk prediction, and how does population attributable fraction quantify the contribution of systemic inflammatory indices? How does machine learning improve mortality risk prediction, and how does population attributable fraction quantify the contribution of systemic inflammatory indices? What is new?Systemic inflammatory indices are independently associated with mortality in DM or PreDM.Machine learning identifies systemic inflammatory indices as key predictors of mortality.Improving systemic inflammatory indices could reduce mortality events by approximately 10–20%. Systemic inflammatory indices are independently associated with mortality in DM or PreDM. Machine learning identifies systemic inflammatory indices as key predictors of mortality. Improving systemic inflammatory indices could reduce mortality events by approximately 10–20%. How might this study influence clinical practice?Integrating systemic inflammatory indices improves risk stratification, and anti-inflammatory strategies may reduce mortality. Integrating systemic inflammatory indices improves risk stratification, and anti-inflammatory strategies may reduce mortality.

慢性全身性炎症是糖尿病（Diabetes Mellitus, DM）及糖尿病前期（prediabetes, PreDM）患者发生心血管代谢并发症的关键诱因。包括中性粒细胞与淋巴细胞比值（neutrophil-to-lymphocyte ratio, NLR）、单核细胞与淋巴细胞比值（monocyte-to-lymphocyte ratio, MLR）、全身炎症反应指数（systemic inflammation response index, SIRI）、全身免疫炎症指数（systemic immune-inflammation index, SII）、血小板与血红蛋白比值（platelet-to-hemoglobin ratio, PHR）及综合全身炎症指数（aggregate inflammation systemic index, AISI）在内的复合炎症指标，已被证实对死亡具有预后评估价值。然而，目前针对这些指标的机器学习整合评估及人群层面的量化研究仍较为有限。
本回顾性队列研究纳入了2005-2018年美国国家健康与营养调查（National Health and Nutrition Examination Survey, NHANES）中的11304名DM或PreDM成人受试者进行分析。本研究的主要结局为全因死亡率与心血管死亡率。采用Cox比例风险模型（Cox proportional hazards models）评估炎症指标与死亡率之间的关联；通过极端梯度提升（Extreme Gradient Boosting, XGBoost）模型评估预测性能，并借助人群归因分数（population attributable fractions, PAFs）估算与全身性炎症相关的死亡负担。
分析结果显示，NLR、MLR、SIRI、SII及AISI均与全因死亡率和心血管死亡率独立相关；其中MLR的关联强度最高（全因死亡率与心血管死亡率的风险比分别为2.948和3.717）。XGBoost模型将SIRI、SII、AISI、MLR及NLR识别为关键预测因子，其中SIRI对心血管死亡率的预测优先级最高。纳入炎症指标可提升模型的区分度与校准度。PAF分析提示，改善全身性炎症状态可降低10%~20%的死亡事件。
全身性炎症指标是DM或PreDM患者死亡率的独立预测因子。将其整合至机器学习模型可优化风险预测效果，可为人群层面的心血管代谢风险分层策略提供参考依据。
目前该领域的已知研究现状为：慢性炎症参与DM或PreDM的病情进展；传统炎症标志物检测成本高昂且难以广泛开展筛查；而全身性炎症指标则为诊断与预后评估提供了一种易于获取的替代方案。慢性炎症参与DM或PreDM的病情进展；传统炎症标志物检测成本高昂且难以广泛开展筛查；而全身性炎症指标则为诊断与预后评估提供了一种易于获取的替代方案。
本研究的核心科学问题为：机器学习如何优化死亡风险预测？人群归因分数如何量化全身性炎症指标对死亡的贡献？机器学习如何优化死亡风险预测？人群归因分数如何量化全身性炎症指标对死亡的贡献？
本研究的创新性发现包括：全身性炎症指标与DM或PreDM患者的死亡率独立相关；机器学习可识别出全身性炎症指标作为死亡的关键预测因子；改善全身性炎症状态可降低约10%~20%的死亡事件。全身性炎症指标与DM或PreDM患者的死亡率独立相关；机器学习可识别出全身性炎症指标作为死亡的关键预测因子；改善全身性炎症状态可降低约10%~20%的死亡事件。
本研究对临床实践的启示为：整合全身性炎症指标可优化风险分层策略，抗炎治疗或可降低患者的死亡风险。整合全身性炎症指标可优化风险分层策略，抗炎治疗或可降低患者的死亡风险。