慢阻肺病是一种全球疾病负担较重的慢性气道疾病，往往累及多个器官，但人们往往聚焦于心、脑等器官，却常常忽视对于肾脏的影响。肺肾已被证实存在密切的交互机制：慢阻肺病以炎症为核心致病通路，肿瘤坏死因子-α（TNF-α）、白细胞介素-6（IL-6）等炎症因子可经由血液循环损伤肾脏，激活NF-κB通路，诱发氧化应激反应，导致肾小管凋亡与间质纤维化；长期处于低氧与高碳酸血症状态，会激活患者体内的低氧诱导因子、收缩肾血管、加重酸中毒，形成肾损伤恶性循环。吸烟不仅仅损伤肺脏，还可通过内皮毒性、氧化应激及镉等重金属蓄积直接损害肾单位。治疗药物的使用及机械通气等治疗措施也可能增加肾损伤风险。传统肾功能指标，例如血肌酐、尿素氮，二者的敏感性及特异性都较低，难以早期发现肾损伤。相较而言，新型生物标志物中，肾损伤分子-1（KIM-1）、中性粒细胞明胶酶相关脂质运载蛋白（NGAL）、尿肝型脂肪酸结合蛋白（L-FABP）可快速提示肾小管的损伤；半胱氨酸蛋白酶抑制剂C（CysC）、β₂-微球蛋白（β₂-MG）能灵敏反映肾小球滤过功能的变化。不仅如此，基于外周血炎症指标构建的白细胞相关炎症指数模型简便易行，适合临床推广；以临床常见指标为基础构建的机器学习模型对慢阻肺病患者急性肾损伤预测效能优异（AUC达0.91)。因此，系统梳理肺-肾交互串扰机制、探索新型生物标志物及相关模型构建，推动人工智能预警，对慢阻肺病相关肾损伤的早期筛查、精准干预与预后改善具有重要的临床价值。

Chronic obstructive pulmonary disease (COPD) has been established as a systemic disorder affecting multiple organ systems, yet its impact on renal function remains frequently underrecognized. A well-documented bidirectional crosstalk exists between the lungs and kidneys. In COPD, chronic inflammation serves as the central pathogenic driver: proinflammatory mediators-including tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6)-enter systemic circulation, directly impair renal tissue, activate the NF-κB signaling pathway, and trigger oxidative stress responses. These processes promote renal tubular apoptosis and interstitial fibrosis. Furthermore, persistent hypoxemia and hypercapnia activate endogenous hypoxia-inducible factors, leading to renal vasoconstriction and severe metabolic acidosis, thereby establishing a self-perpetuating cycle of progressive kidney injury. Cigarette smoking exerts dual organ toxicity: beyond its well-known pulmonary effects, it induces direct nephrotoxicity via endothelial dysfunction, oxidative stress, and accumulation of heavy metals such as cadmium. Additionally, therapeutic interventions-including pharmacologic agents and mechanical ventilation-may further elevate the risk of acute r chronic kidney injury in COPD patients. Conventional renal biomarkers, such as serum creatinine and blood urea nitrogen, exhibit limited sensitivity and specificity for early detection of kidney damage, often failing to identify subclinical injury until significant functional decline has occurred. In contrast, novel biomarkers offer superior diagnostic performance: kidney injury molecule-1 (KIM-1), neutrophil gelatinase-associated lipocalin (NGAL), and urinary liver-type fatty acid-binding protein (L-FABP) provide rapid, sensitive indicators of proximal tubular injury, whereas cystatin C (CysC) and β-microglobulin (β-MG) reliably reflect subtle alterations in glomerular filtration rate. Moreover, the leukocyte-related inflammatory index model-constructed using readily accessible peripheral blood inflammatory markers-is simple, practical, and highly suitable for clinical implementation. Furthermore, a machine learning model built upon routinely available clinical parameters demonstrates excellent predictive performance for acute kidney injury in COPD patients, achieving an area under the receiver operating characteristic curve of 0.91. Therefore, systematically elucidating the bidirectional crosstalk mechanisms between the lung and kidney, identifying novel biomarkers, and developing associated predictive models-particularly AI-driven early warning systems-hold significant clinical value for the early screening, precise intervention, and improved prognosis of COPD-associated renal injury.