We developed models for individualized risk prediction of cognitive decline in mild cognitive impairment (MCI), using plasma biomarkers of beta-amyloid (Ab), tau, and neurodegeneration. We included MCI patients from the Swedish BioFINDER study (n=148) and the Alzheimer's Disease Neuroimaging Initiative (ADNI; n=86 for model selection; n=425 for prognostic validation). The primary outcomes were longitudinal cognition and conversion to AD dementia, predicted by plasma Ab42/Ab40, P-tau181, and neurofilament light (NfL). A model which included P-tau181 and NfL, but not Ab42/Ab40, had the best performance (AUC=0.88 for four-year conversion to AD in BioFINDER, validated in ADNI). The prognostic ability of plasma biomarkers was stronger than a basic model of age, sex, education, and baseline cognition and similar to cerebrospinal fluid biomarkers. Plasma biomarkers, in particular P-tau181 and NfL, may be of high value to identify MCI individuals who will progress to AD dementia in clinical trials and in clinical practice.

中文翻译：

---

血浆淀粉样蛋白，磷酸化tau蛋白和神经丝光可用于轻度认知障碍的个体化风险预测

我们使用β-淀粉样蛋白（Ab），tau和神经退行性变的血浆生物标记物开发了用于轻度认知障碍（MCI）认知下降的个体化风险预测模型。我们纳入了来自瑞典BioFINDER研究（n = 148）和阿尔茨海默氏病神经影像学计划（ADNI；模型选择为n = 86；预后验证为n = 425）的MCI患者。主要结果是通过血浆Ab42 / Ab40，P-tau181和神经丝轻（NfL）预测的纵向认知并转变为AD痴呆。包含P-tau181和NfL但不包含Ab42 / Ab40的模型具有最佳性能（在BioFINDER中四年转换为AD的AUC = 0.88，在ADNI中进行了验证）。血浆生物标志物的预后能力强于年龄，性别，教育程度和基线认知的基本模型，并且与脑脊液生物标志物相似。