目的： 铁是人体必需的微量元素。有大量研究表明全血捐献会引起铁的流失。随着单 采血浆事业的发展，单采血浆捐献者的健康越来越受到关注。然而单采血浆捐献者 铁代谢状况我们却知之甚少。本研究旨在探究单采血浆捐献对铁代谢的影响，筛查 发现铁缺乏风险人群及其影响因素，并进一步通过建立与验证铁缺乏风险预测模型， 实现风险人群早期预警和干预，从而保护献浆者健康安全。

第一部分：单采血浆捐献对铁代谢影响研究

方法： 从湖南、云南、河南、四川四省五家单采血浆站招募单采血浆捐献者。招募新 献浆者为对照组，近一年有献浆行为的献浆者（重复献浆者）为调查组。检测两组 人群的血清铁蛋白（serum ferritin，SF）和血红蛋白（Hemoglobin，Hb）。采用多 重线性回归校正混杂因素后探究单采血浆捐献对 SF 和 Hb 检测值的影响。采用多因 素 Logistic 回归校正混杂因素后探究单采血浆捐献对铁代谢异常率的影响。 结果： 共纳入 726 名新献浆者和 767 名重复献浆者。校正混杂因素后，高献浆频次组 （献浆次数>16 次/年）献浆者 SF 水平均显著低于新献浆者（男性： P<0.001；女 性： P=0.008）。女性重复献浆者 Hb 水平显著低于新献浆者（P<0.001），男性重复 献浆者 Hb 水平与新献浆者无显著差异。男性重复献浆者铁蓄积率为 38.97%，女性 重复献浆者的铁蓄积率为 8.49%，高献浆频次及高献浆总次数组男性献浆者铁蓄积 率显著低于新献浆者（高献浆频次组 vs.新献浆者：P=0.001；高献浆总次数组 vs. 新献浆者：P<0.001）。男性重复献浆者铁缺乏率为 0.78%，女性重复献浆者的铁缺 乏率为 28.38%，且高献浆频次组女性重复献浆者的铁缺乏率明显高于新献浆者 （P=0.01)。 结论： 在中国现行的单采血浆捐献标准下，较高的献浆频次及献浆总次数会在一定程 度上降低捐献者 SF。单采血浆捐献在一定程度上可降低男性捐献者铁蓄积的风险， 但也会在一定程度上增加女性献浆者铁缺乏的风险。 

第二部分：女性重复献浆者铁缺乏风险因素分析及预测模型的建立与验证

方法： 应用多因素 Logistic 回归分析探究女性重复献浆者铁缺乏独立风险因素。将 女性重复献浆者按 7： 3 的比例随机分配到建模和验证队列中。将独立风险因素纳 入多因素 Logistic 回归模型中，用逐步回归法进行预测模型构建。招募另一地区 女性重复献浆者作为外部验证队列。采用受试者工作特征曲线和（receive operator characteristic curve, ROC）和 Hosmer-Lemeshow 拟合优度检验对构建的预测模 型在建模及验证队列中的区分度和校准度进行评价，采用临床决策曲线来评价模型 的临床价值。 结果： 由于女性重复献浆者的铁缺乏率较高，我们对女性重复献浆者的铁缺乏进行了 风险因素分析，结果显示，捐献频次、年龄、Hb 检测值、身体质量指数（Body Mass Index，BMI）和绝经前女性是铁缺乏的独立风险因素（P<0.05）。基于独立风险因素 分析结果建立并验证了女性重复献浆者铁缺乏风险预测模型。预测模型在建模队列、 内部验证队列以及外部验证队列中均具有良好的区分度（AUC>0.7）和校准度 （P>0.05），模型预测风险与实际发生风险的一致程度较好，模型预测准确率较高。 结论： 年龄、BMI、献浆频次、Hb 检测值以及月经史与女性重复献浆者铁缺乏的发生 高度相关，月经史是女性重复献浆者铁缺乏最重要的预测因素。其中，年龄、BMI、 Hb 检测值与铁缺乏风险呈负相关；献浆频次与铁缺乏风险呈正相关；未绝经女性铁 缺乏风险较高。建立的铁缺乏风险预测模型具有较好的区分度与校准度，具备一定 的预测能力。 

Objective： Iron is an essential trace element for the human body. Numerous studies have shown that whole blood donation can cause iron loss. With the development of the plasmapheresis donation, the health of single plasma donors is receiving increasing attention. However, we know very little about the iron metabolism status of single plasma donors. The aim of this study is to explore the impact of plasma donation on iron metabolism, screen and identify individuals at risk of iron deficiency and their influencing factors, and further establish and validate an iron deficiency risk prediction model to achieve early warning and intervention for high-risk individuals, thereby protecting the health and safety of plasma donors.

Section 1: Study on the Effect of Plasmapheresis Donation on Iron Metabolism

Methods: Recruit single plasma donors from five single plasma collection stations in Hunan, Yunnan, Henan, and Sichuan provinces. Recruit new plasma donors as the control group, and plasma donors (repeat donors) who have exhibited plasma donation behavior in the past year as the investigation group. Detect serum ferritin (SF) and hemoglobin (Hb) in two groups of people. Exploring the impact of plasma donation on SF and Hb detection values after adjusting for confounding factors using multiple linear regression. Exploring the effect of plasma donation alone on the incidence of iron metabolism abnormalities after adjusting for confounding factors using multivariate logistic regression. Results: A total of 726 new donors and 767 repeat donors were included. After adjusting for confounding factors, the SF levels of plasma donors in the high plasma donation frequency group (plasma donation frequency>16 times/year) were significantly lower than those in the new plasma donors (male: P<0.001; female: P=0.008). The Hb levels of female repeat plasma donors were significantly lower than those of new plasma donors (P<0.001), while there was no significant difference in Hb levels between male repeat plasma donors and new plasma donors. The iron accumulation rate of male repeated plasma donors was 38.97%, while that of female repeated plasma donors was 8.49%. The iron accumulation rate of male plasma donors with high plasma frequency and high total plasma frequency arrays was significantly lower than that of new plasma donors (high plasma frequency group vs. new plasma donors: P=0.001; high total plasma frequency array vs. new plasma donors: P<0.001). The iron deficiency rate of male repeat plasma donors was 0.78%, while that of female repeat plasma donors was 28.38%. The iron deficiency rate of female repeat plasma donors in the high frequency group was significantly higher than that of new plasma donors (P=0.01). Conclusions: Under the current single plasma donation standards in China, a higher frequency and total number of plasma donations will to some extent reduce donor SF. Single plasma donation can to some extent reduce the risk of iron accumulation in male donors, but it also increases the risk of iron deficiency in female plasma donors. 

Section 2: Analysis of Risk Factors for Iron Deficiency in Female Repeated Plasma Donors and Development and Validation of Predictive models

Method: Applying multivariate logistic regression analysis to explore independent risk factors for iron deficiency in female repeat plasma donors. Randomly assign female repeat plasma donors in a 7:3 ratio to the modeling and validation queue. Incorporate independent risk factors into a multivariate logistic regression model and construct a predictive model using stepwise regression forward backward method. Recruit female repeat plasma donors from another region as an external validation queue. The receiver operator characteristic curve (ROC) and Hosmer-Lemeshow Goodness of fit test were used to evaluate the discrimination and calibration of the constructed prediction model in the modeling and validation queue, and the clinical decision curve was used to evaluate the clinical value of the model. Result: Due to the high iron deficiency rate among female repeat plasma donors, we conducted a risk factor analysis on iron deficiency among female repeat plasma donors. The results showed that donation frequency, age, Hb test value, Body Mass Index (BMI), and premenopausal women were independent risk factors for iron deficiency (P<0.05). A risk prediction model for iron deficiency in female repeat plasma donors was established and validated based on the results of independent risk factor analysis. The prediction model has good discrimination (AUC>0.7) and calibration (P>0.05) in the modeling queue, internal and validation queues, and external validation queues. The model has a good consistency between predicted risks and actual risks, and the model has a high prediction accuracy. Conclusion: Age, BMI, frequency of plasma donation, Hb detection value, and menstrual history are highly correlated with the occurrence of iron deficiency in female repeat plasma donors. Menstrual history is the most important predictor of iron deficiency in female repeat plasma donors. Among them, age, BMI, and Hb detection values are negatively correlated with the risk of iron deficiency; The frequency of plasma donation is positively correlated with the risk of iron deficiency; Premenopausal women have a higher risk of iron deficiency. The established iron deficiency risk prediction model has good discrimination and calibration, and has certain predictive ability.