Yang Zhixun, van Os Hendrikus J A, Kist Janet M, Vos Rimke C, Vos Hedwig M M, Chavannes Niels H, Petrus Annelieke H J
Department of Public Health & Primary Care, Leiden University Medical Center, Albinusdreef 2, Leiden, 2333, ZA, the Netherlands.
Health Campus The Hague, Leiden University Medical Center, Albinusdreef 2, Leiden, 2333, ZA, the Netherlands.
Int J Cardiol Cardiovasc Risk Prev. 2025 Aug 5;27:200483. doi: 10.1016/j.ijcrp.2025.200483. eCollection 2025 Dec.
Pregnancy-related factors are associated with an increased risk of cardiovascular disease (CVD) and may help identify women at high cardiovascular risk. This study aims to provide an overview of prediction models for CVD which included pregnancy-related factors and to evaluate the impact of these factors on model performance.
PubMed and Embase were systematically searched until March 2023 for studies reporting on the development or validation of prediction models for CVD which included pregnancy-related factors. Data extraction was performed using the CHARMS checklist. Risk of bias was assessed using PROBAST.
Seven studies were included. C-indices ranged between 0.63 and 0.79. Adding pregnancy-related factors resulted in improved C-index in four studies, ranging from 0.0033 (95 % confidence interval [CI]: 0.0022-0.0051) to 0.004 (95 % CI: 0.002-0.006). Net reclassification improvement (NRI) for events was improved in two studies, ranging from 0.01 (95 % CI: 0.003-0.02) to 0.038 (95 % CI: 0.003-0.074). NRI for non-events was improved in three studies, ranging from 0.002 (95 % CI: 0.0001-0.005) to 0.02 (95 % CI: 0.001-0.04). Two studies showed both low risk of bias and low concern regarding applicability. Subgroup analyses by age in three studies indicated larger improvements in model performance in younger women.
Addition of pregnancy-related factors results in limited improvements in performance of CVD prediction models, with relatively larger improvements in younger women.
与妊娠相关的因素与心血管疾病（CVD）风险增加相关，可能有助于识别心血管疾病高风险女性。本研究旨在概述包含与妊娠相关因素的心血管疾病预测模型，并评估这些因素对模型性能的影响。
系统检索PubMed和Embase直至2023年3月，以查找报告包含与妊娠相关因素的心血管疾病预测模型开发或验证的研究。使用CHARM清单进行数据提取。使用PROBAST评估偏倚风险。
纳入七项研究。C指数在0.63至0.79之间。在四项研究中，添加与妊娠相关的因素使C指数得到改善，改善幅度从0.0033（95%置信区间[CI]：0.0022 - 0.0051）至0.004（95%CI：0.002 - 0.006）。两项研究中事件的净重新分类改善（NRI）得到改善，范围从0.01（95%CI：0.003 - 0.02）至0.038（95%CI：0.003 - 0.074）。三项研究中非事件的NRI得到改善，范围从0.002（95%CI：0.0001 - 0.005）至0.02（95%CI：0.001 - 0.04）。两项研究显示偏倚风险低且适用性方面担忧少。三项研究按年龄进行的亚组分析表明，年轻女性的模型性能改善更大。
添加与妊娠相关的因素使心血管疾病预测模型的性能改善有限，年轻女性的改善相对更大。
Zotero 插件