BACKGROUND Soluble triggering receptor expressed on myeloid cells 2 (sTREM2) plays an important role in the clearance of pathological amyloid-β (Aβ) in Alzheimer’s disease (AD). This study aimed to explore sTREM2 as a central and peripheral predictor of the conversion from mild cognitive impairment (MCI) to AD. METHODS sTREM2 and Aβ 1–42 levels in cerebrospinal fluid (CSF) and florbetapir-PET (AV45) images were analyzed for healthy control (HCs), patients with MCI, and patients with AD from the ADNI database. Peripheral plasma sTREM2 and Aβ 1–42 levels were determined for our Neurology database of Ruijin Hospital for Alzheimer’s Disease (NRHAD) cohort, and patients with MCI were reevaluated at follow-up visits to assess for progression to AD. The association between CSF and plasma sTREM2 levels was analyzed in data from the Chinese Alzheimer’s Biomarker and Lifestyle (CABLE) database. RESULTS The results showed that patients with MCI who had low levels of CSF sTREM2 and Aβ 1–42 were more likely to develop AD. Among participants with positive Aβ deposition, as assessed by AV45 imaging, elevated CSF sTREM2 levels were associated with a decreased risk of MCI-to-AD conversion. Meanwhile, in the NRHAD cohort, individuals in the MCI group with high sTREM2 levels in plasma were at a greater risk for AD, whereas low Aβ 1–42 with high sTREM2 levels in plasma were associated with a faster cognitive decline. In addition, CSF sTREM2 levels were highly correlated with plasma sTREM2 levels in the CABLE database. CONCLUSION These findings suggest that sTREM2 may be useful as a potential predictive biomarker of MCI-to-AD conversion. FUNDING This study was supported by grants from the National Natural Science Foundation of China (grant nos. 82001341, 82071415, 81873778, and 82201392); the Shanghai Sailing Program (grant no. 22YF1425100); and the China Postdoctoral Science Foundation funded project (grant no. 2021M702169).

In some cases, near-infrared spectra (NIRS) make the prediction of quantitative models unreliable, and the choice of a suitable number of latent variables (LVs) for partial least square (PLS) is difficult. In this case, a strategy of fusing member models with important information is gradually becoming valued in recent research. In this work, a series of PLS regression models were developed with an increasing number of LVs as member models. Then, the least absolute shrinkage and selection operator (Lasso) was employed as the model’s selection access to sparse uninformative ones among these PLS member models. Deviation weighted fusion (DW-F), partial least squares regression coefficient fusion (PLS-F), and ridge regression coefficient fusion (RR-F) were comparatively used further to fuse the above sparsed member models, respectively. Three spectral datasets, including six attributes in NIR data of corn, apple, and marzipan, respectively, were applied in order to validate the feasibility of this fusion algorithm. Six fusion models of the above attributes performed better than the general optimal PLS model, with a noticeable enhancement of root mean errors squared of prediction (RMSEP) arriving at its highest at 80%. It also reduced more than half of the spectral bands; the DW-F especially showed its excellent fusing capacity and obtained the best performance. Results show that the preferred strategy of DW-F model combined with Lasso selection can make full use of spectral information, and significantly improve the prediction accuracy of fusion models.