Among patients with isolated AD or isolated CSVD, plasma p-Tau217 is the optimal specific biomarker for the diagnosis of AD and differentiation between AD and cerebral small vessel disease, while GFAP acts as a key indicator for identifying CSVD.
Key Findings
Results
Plasma Aβ1-42 and Aβ1-42/Aβ1-40 ratio were significantly decreased in the AD group compared to CSVD and healthy controls.
120 participants were enrolled: 40 in the AD group, 40 in the CSVD group, and 40 in the healthy control group.
Plasma biomarkers were detected by chemiluminescence immunoassay.
The decrease in Aβ1-42 and Aβ1-42/Aβ1-40 ratio was described as 'significantly decreased' in the AD group.
Cognitive function and neuroimaging examinations were completed simultaneously with biomarker detection.
Results
Plasma p-Tau181 and p-Tau217 were markedly elevated in the AD group compared to CSVD and healthy controls.
Both p-Tau181 and p-Tau217 were described as 'markedly elevated' in the AD group.
The elevation was observed relative to both the CSVD group and the healthy control group.
Each group contained 40 participants for comparison.
Results
GFAP levels in the CSVD group were specifically and significantly higher than in both the AD group and healthy control group.
GFAP elevation was described as 'specifically and significantly higher' in CSVD compared to both other groups.
This pattern distinguished CSVD from AD, as GFAP was not similarly elevated in the AD group.
GFAP showed an AUC of 0.881 for distinguishing CSVD from healthy controls.
GFAP was identified as 'a key indicator for identifying CSVD.'
Results
NfL was significantly increased in both the AD and CSVD disease groups compared to healthy controls.
NfL elevation was not specific to either disease group, being elevated in both AD and CSVD.
This non-specificity suggests NfL reflects general neurodegeneration rather than disease-specific pathology.
NfL was among the seven biomarkers evaluated: Aβ1-42, Aβ1-40, Aβ1-42/Aβ1-40 ratio, p-Tau181, p-Tau217, NfL, and GFAP.
Results
p-Tau217 demonstrated optimal efficacy for distinguishing AD from healthy controls with an AUC of 0.894.
ROC curve analysis was used to evaluate diagnostic efficacy.
AUC for p-Tau217 distinguishing AD from healthy controls was 0.894.
This was the highest AUC among single biomarkers for AD versus healthy control discrimination.
p-Tau217 was identified as 'the optimal specific biomarker for the diagnosis of AD.'
Results
p-Tau217 showed the best performance for differentiating AD from CSVD with an AUC of 0.877.
AUC for p-Tau217 in differentiating AD from CSVD was 0.877.
This was the highest AUC among single biomarkers for the AD versus CSVD differential diagnosis.
Spearman correlation analysis was used alongside ROC analysis to assess biomarker relationships.
p-Tau217 was described as optimal for 'differentiation between AD and cerebral small vessel disease.'
Results
All seven plasma biomarkers were significantly correlated with MMSE scores.
Spearman correlation analysis was used to analyze the correlation between each biomarker and MMSE score.
All biomarkers — Aβ1-42, Aβ1-40, Aβ1-42/Aβ1-40 ratio, p-Tau181, p-Tau217, NfL, and GFAP — showed significant correlations with MMSE scores.
MMSE was the cognitive assessment tool used across all 120 participants.
Discussion
The diagnostic utility of these biomarkers should be interpreted cautiously in patients with AD-CSVD co-pathology.
The authors noted that findings 'should be interpreted cautiously for patients with AD-CSVD co-pathology.'
The study population consisted of pathologically isolated cases (isolated AD or isolated CSVD), which may limit generalizability.
Combined detection of multiple plasma biomarkers was recommended for 'non-invasive early screening and etiological classification of cognitive impairment.'
What This Means
This research suggests that specific proteins measurable in blood can help doctors distinguish between two common causes of memory and thinking problems in older adults: Alzheimer's disease (AD) and cerebral small vessel disease (CSVD). The study tested seven different blood-based markers in 120 people divided equally into three groups — those with AD, those with CSVD, and healthy adults. The researchers found that different diseases leave distinct 'fingerprints' in the blood: AD was associated with lower levels of amyloid proteins (Aβ1-42) and higher levels of tau proteins (especially p-Tau217), while CSVD was uniquely associated with elevated GFAP, a marker of brain cell damage. A protein called NfL was elevated in both diseases, making it less useful for telling them apart.
The most practically significant finding was that p-Tau217 performed best at identifying AD (AUC = 0.894, meaning about 89% accuracy) and at distinguishing AD from CSVD (AUC = 0.877). GFAP was the best marker for identifying CSVD from healthy individuals (AUC = 0.881). All markers showed meaningful correlations with standard cognitive test scores (MMSE), suggesting they reflect the degree of cognitive impairment.
This research suggests that simple blood tests — rather than expensive or invasive brain scans and spinal fluid procedures — could help clinicians identify which type of brain disease is causing a patient's cognitive decline. However, the authors caution that these results apply most clearly to patients who have only one of these conditions. Many real-world patients have both AD and CSVD simultaneously, and the markers' usefulness in that mixed group remains uncertain. The authors recommend using multiple biomarkers together rather than relying on any single one for clinical decision-making.
Xiandong W, Junqi W, Xuan Z, Daichao M, Xiaoyan Z, Lihua H, et al.. (2026). Expression differences and diagnostic efficacy of core plasma biomarkers in Alzheimer's disease, cerebral small vessel disease and healthy adults.. Frontiers in neurology. https://doi.org/10.3389/fneur.2026.1934127