Neuropathological links between plasma p-Tau 181, white matter hyperintensity, and structural brain changes in aging

Introduction

White matter hyperintensities (WMHs) are commonly observed in individuals aged over 60 years, and their prevalence increases with age (1). Conventionally, WMHs are regarded as a hallmark of cerebral small vessel disease (CSVD), and are associated with an increased risk of stroke, cognitive decline, and depression (2). However, recent research suggests that WMHs may be directly linked to the intrinsic pathological processes of Alzheimer’s disease (AD), driven by AD-specific pathologies, including amyloid-beta (Aβ) deposition and tauopathy (3,4). For instance, the spatial distribution patterns of WMH in AD often overlap with regions exhibiting hypometabolism and neurodegeneration.

Reductions in brain volume, cortical thinning, white matter degeneration, gyral simplification, and ventricular enlargement represent the most prominent morphological features of aging (5). Age-related atrophy is particularly pronounced in the medial temporal and posterior regions, including the hippocampus, and this pattern of structural loss closely resembles that observed in AD, although it progresses more slowly (6). Even among cognitively healthy adults, such changes have been shown to predict subsequent memory decline (7). In non-demented middle-aged and older individuals, reduced cortical thickness has also been reported to partially mediate the association between advancing age and diminished functional connectivity in specific neural subnetworks (8).

Due to their low invasiveness and potential for predicting cognitive decline and conversion to AD, plasma biomarkers such as phosphorylated tau (p-Tau), Aβ, and neurofilament light chain have recently attracted significant research attention. Plasma p-Tau 181 levels exhibit an age-related increase across individuals aged 30 to 98 years, and have been shown to predict amyloid pathology abnormalities, such as Aβ-positron emission tomography (PET) positivity (9). This age-associated increase has also been observed in cognitively normal individuals (10). Moreover, elevated p-Tau 181 concentrations are significantly associated with multidomain cognitive decline, particularly in individuals with mild cognitive impairment or carriers of the apolipoprotein E (APOE) ε4 allele (11). Importantly, research has shown that plasma p-Tau 181 can be used to distinguish between AD and primary age-related tauopathy (12). Although plasma p-Tau 181 is closely associated with age-related cognitive decline and AD pathology, its specificity may be influenced by coexisting pathologies.

Plasma p-Tau 181, WMH, and structural brain alterations interact in a complex manner during aging, jointly driving neurodegenerative processes (13). Higher plasma p-Tau 181 levels have been reported to be significantly correlated with increasing WMH volumes in patients with cognitive impairment (14,15). Further, higher concentrations of plasma p-Tau 181 have been found to be negatively correlated with gray and white matter volume in aging, both cross-sectionally and longitudinally (16). In addition, the WMH burden has been linked to cognitive decline via its effect on cortical and medial temporal lobe integrity, and reductions in WMH over time have been shown to be associated with less brain atrophy and improved memory performance (17,18).

To date, most existing studies have examined associations between any two of the following factors: plasma biomarkers, WMH, and brain morphology. However, research addressing all three factors in combination remains limited, particularly among cognitively normal older adults, in whom these factors are commonly observed. Therefore, this study aimed to explore the potential link that simultaneously integrates these three interrelated factors in older adults, providing insights into how cerebrovascular and AD-related pathology jointly contribute to brain aging. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-376/rc).

Methods

Participant population

In total, 582 elderly individuals were recruited from the Outpatient Clinic of the Departments of Neurology and Geriatrics at Nanjing Drum Tower Hospital between September 2020 and December 2021, and underwent plasma p-Tau 181 and Aβ42 testing to study brain aging. The inclusion criteria were as follows: (I) age ≥50 years; (II) undergoing a health examination, or presenting with dizziness or sleep disturbances; (III) presence or absence of vascular risk factors, including grade 1 hypertension, type 2 diabetes, or prior lacunar infarction (LI); (IV) availability of complete brain magnetic resonance imaging (MRI) data; and (V) right-handedness. The exclusion criteria were as follows: (I) neurodegenerative disorders; (II) a history of ischemic stroke with infarctions >1.5 cm in diameter; (III) extracranial or intracranial large artery stenosis >50%; and/or (IV) intracranial hemorrhage. Details of the data screening workflow are shown in Figure 1. Ultimately, 218 individuals were included in the final analysis. It should be noted that when comparing the differences in brain structure between the different WMH severity groups, only three-dimensional (3D) T₁-weighted (T₁W) images were used. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Nanjing Drum Tower Hospital (IRB review approval No. 2022-165-01). Informed consent was obtained from all the participants.

Figure 1 Flowchart of study enrollment. 2D, two-dimensional; 3D, three-dimensional; Aβ42, amyloid-beta 42; AD, Alzheimer’s disease; CSVD, cerebral small vessel disease; DWI, diffusion-weighted imaging; FLAIR, fluid-attenuated inversion recovery; MRI, magnetic resonance imaging; p-Tau, phosphorylated tau; T₁WI, T1-weighted imaging; T₂WI, T2-weighted imaging; WMH, white matter hyperintensity.

MRI data acquisition

The MRI data of the participants were collected using a 3.0-Tesla MR scanner (Achieva 3.0T TX dual Medical Systems; Philips Medical Systems, Eindhoven, Netherlands). Various sequences, such as T₁W, T2-weighted (T₂W), T₂ fluid-attenuated inversion recovery (FLAIR), and diffusion-weighted imaging (DWI) images, were obtained to detect CSVD-related lesions and other brain abnormalities. The structural imaging data were either in 2D or 3D format. The following parameters were used for the 2D T₁W diagnostic images: repetition time (TR) =110 ms, echo time (TE) =1,580 ms, flip angle =8°, slice thickness =6 mm, slice gap =7 mm, and field of view (FOV) =256 mm × 256 mm. The following parameters were used for the T₂W images: TR =3,908 ms, TE =100 ms, flip angle =90°, and FOV =288 mm × 288 mm. The following parameters were used for the FLAIR images: TR =5,600 ms, TE =115 ms, flip angle =90°, and FOV =256 mm × 256 mm. The following parameters were used for the 3D T₁W isotropic anatomical images: TR =6,500 ms, TE =300 ms, flip angle =8°, slice thickness =1 mm, slice gap =1 mm, and FOV =240 mm × 240 mm. The following parameters were used for the DWI images: TR =2,988 ms, TE =90 ms, flip angle =90°, FOV =192 mm × 192 mm, slice thickness =5 mm, and slice gap =6 mm.

Analysis of CSVD types

Five types of imaging biomarkers were identified based on the 2021 Chinese expert consensus on the diagnosis and treatment of CSVD. A radiologist with 5 years of experience performed the initial review of the MRI images, assessing and recording the presence of recent small subcortical infarction (RSSI), LI, perivascular space (PVS) and cerebral atrophy (CA). The criteria for these imaging markers are described in a recent review (19). The images were then examined by another radiologist. A third radiologist made the final decision if any disagreements arose.

The Fazekas scale was used to evaluate the extent of WMH on a scale from 0 to 3, where 0 represents no lesions; grade 1 represents a few small isolated foci of hyperintensity; grade 2 represents a moderate confluence of small lesions, forming larger areas of hyperintensity; and grade 3 represents an extensive confluence of lesions, resulting in large areas of WMH. Based on WMH severity, all the participants were allocated to the mild WMH group (Fazekas scores of 0 and 1) or severe WMH group (Fazekas scores of 2 and 3).

Plasma Aβ42 and p-Tau 181 measurement

Enzyme-linked immunosorbent assay (ELISA) was used to measure the plasma biomarkers. After a period of fasting, 5 mL of peripheral venous blood was collected in ethylenediaminetetraacetic acid anticoagulant tubes from the participants. The blood samples were centrifuged at 2,000 ×g for 15 minutes within an hour of collection. Subsequently, the plasma samples were stored at −80 ℃ to avoid any repeated freeze-thaw cycles. The time interval from sampling to testing ranged from 16 to 379 days. Plasma sample concentrations were determined using a Varioskan LUX multimode microplate reader (Thermo Scientific, USA). The human β-amyloid protein 1-42 (Cat. No. 290-62601, Shenzhen AnQun Biological Engineering Co., Ltd., Shenzhen, China) and human p-Tau-181 protein assay kits (Cat. No. 298-81701, Shenzhen AnQun Biological Engineering Co., Ltd.) were used to measure the amyloid biomarkers (Aβ42) and tau biomarkers (p-Tau 181). The assays were conducted by personnel who were blinded to the participants’ information.

Pre-processing of 3D T₁W images

FreeSurfer version 6.0.0 (https://surfer.nmr.mgh.harvard.edu/fswiki/rel6downloads) was used to pre-process the brain structural images of each participant. The pre-processing was performed using the automated surface-based pipeline with default parameters, including segmentation, surface reconstruction, and surface-based spatial registration. Initially, the structural images were aligned to the Talairach atlas (20), and the normalization of white matter intensity variation was conducted. Subsequently, the skull data were removed, and segmentation was performed to identify three tissue types [white matter, gray matter, and cerebrospinal fluid (CSF)]. Any inaccuracies in segmentation were manually corrected using Freeview (a visualization tool packaged with FreeSurfer) to enhance the quality of the segmentation results. A 10-mm full-width-at-half-maximum Gaussian spatial smoothing kernel was then applied to enhance the signal-to-noise ratio.

Cortical structural alterations between mild and severe WMHs

All the structural images (original 3D T₁W and reconstructed 3D T₁W images) were divided into the left and right hemispheres by FreeSurfer. Cortical parcellation was performed according to Destrieux’s Atlas (2009) (21), which divides each hemisphere into 74 gyral and sulcal regions (148 regions in total) based on standard anatomical nomenclature. All cortical regions from both hemispheres were included in the analyses. For each hemisphere, the multivariate general linear model (GLM) was used based on original 3D T₁W images (n=83), with age and sex as nuisance covariates, to measure the group differences between individuals with mild and severe WMH. False discovery rate (FDR) correction was used for multiple comparison corrections.

Subcortical nuclei alterations between mild and severe WMHs

Subcortical segmentation was performed using the probabilistic atlas implemented in FreeSurfer, which labels major subcortical gray matter nuclei, including the thalamus, caudate, putamen, pallidum, hippocampus, amygdala, and nucleus accumbens. These seven nuclei were included in the statistical analyses. Further, the estimated total intracranial volume (TIV) was extracted to adjust for head size differences. The multivariate GLM was used with age, sex, and TIV as nuisance covariates to measure the group differences in subcortical nuclei volumes between mild and severe WMHs, based on the original 3D T₁W images. FDR correction was used for multiple comparison corrections. Significant nuclei volumes were examined in further correlation analysis and mediation analysis.

Super-resolution processing for diagnostic imaging data

A subset of 2D T₁W images was reconstructed into 3D format for structural analysis. The structure-constrained super-resolution network (SCSRN) was employed. Spatial similarity and regional evaluations showed that SCSRN outperformed other deep learning and interpolation techniques in accuracy and reliability, as validated on 1,611 MRI cases (22). Initially, the input image underwent upsampling via trilinear interpolation to match the resolution of the desired ground-truth image. Subsequently, a ResNet algorithm was employed as the backbone of our network, removing the pooling layer to retain resolution, and implementing a residual scaling factor of 0.1 to stabilize training. Further, the batch normalization layer was omitted to maintain the flexibility of the feature ranges. Finally, a combination of mean square error loss and Dice loss served as the loss function to enhance both image intensity and structural similarity, with a pre-trained segmentation model used to identify brain structures.

Harmonization of magnetic resonance measurements

To amplify the statistical effect, all the T₁W images were included in the final mediation analysis. Combat processing was performed on the original 3D T₁W and reconstructed 3D T₁W images before integration. ComBat, a statistical method designed to correct for batch effects in datasets, was used to harmonize the nuclei volumes across different scanners or imaging protocols. It adjusts the statistical distributions of radiomic feature values obtained from various centers or scanners. ComBat has been shown to be applicable to various magnetic resonance sequences, including structural MRI (23). ComBaTool (24), a free online application (available at https://forlhac.shinyapps.io/Shiny_ComBat/), was also used to conduct the harmonization of the bilateral caudate nucleus volumes segmented from different sources of 3D T₁W images.

Statistical analysis

The Mann-Whitney U test was used to assess group differences in the continuous demographic variables, while the chi-square test was used for the categorical variables. Spearman correlation coefficients were calculated to measure the correlation between plasma biomarkers and age. Differences in correlation between different groups were analyzed using Fisher’s r-to-z transformation method. Additionally, group differences in cortical thickness and nuclei volumes were assessed using a GLM, controlling for age, sex, and TIV. All the statistical analyses were performed using SPSS (version 23.0, IBM Corp.). The association between plasma Aβ42/p-Tau 181 and clinical conditions, as well as the CSVD types, was analyzed using multivariate logistic regression in R (version 4.4.3, R Foundation). Separate models were fitted for each WMH severity group, adjusting for age and sex. Marginal effects were visualized using predicted probabilities from the fitted models. Moreover, partial correlation analyses were performed in R to assess the associations between subcortical brain volumes and plasma biomarkers, controlling for age, sex, and TIV, as well as the plasma biomarkers and WMH burden, adjusting for age and sex as covariates. Partial correlation analyses were conducted to examine the associations between subcortical brain volumes and age or sex. Specifically, when assessing the correlation with age, sex, and TIV were included as covariates, while when assessing the correlation with sex, age and TIV were included as covariates. FDR corrections were applied separately, with a statistical significance level set at P<0.05. Additionally, the interaction between age and WMH severity on caudate nucleus volume was examined using linear regression models that included an interaction term (age × WMH severity), adjusting for sex and TIV. The mediation analysis was conducted using structural equation modeling in R to examine the indirect effect of plasma biomarker levels on caudate nucleus volumes through Fazekas scores, controlling for sex, age, and TIV. The bootstrap sensitivity analysis with 500 replications was employed to estimate confidence intervals (CIs) and assess the stability of the mediation effect.

Results

Demographic and clinical data

The participants were allocated to the mild WMH group (Fazekas scores ≤1) or severe WMH group (Fazekas scores ≥2). There were no statistically significant differences in the prevalence of hypertension or diabetes between the mild and severe WMH groups (hypertension: 7.1% vs. 9.5%, P=0.512; diabetes: 4.4% vs. 7.6%, P=0.320), ruling out any potential confounding effects of these diseases on the study results. As shown in Table 1, the participants in the severe WMH group had a higher age and higher plasma p-Tau 181 levels than those in the mild WMH group. However, no significant differences in gender and plasma Aβ42 were observed between the two groups. Further, the prevalence of CA, RSSI, and LI was elevated in the severe WMH group.

Associations between plasma AD pathological biomarkers and age

In all participants, the plasma p-Tau 181 and Aβ42 levels were positively correlated with age (Figure 2A,2B). Moreover, the plasma p-Tau 181 levels were positively correlated with the plasma Aβ42 concentrations (Figure 2C). Additionally, the correlation between plasma Aβ42 levels and age was higher in the mild WMH group (z=2.04, P=0.042), while the correlation between the plasma p-Tau 181 levels and age was similar in both the mild and severe WMH groups (z=−0.72, P=0.472) (Figure 2D). Further, positive correlations between the plasma biomarkers and age were observed in both males and females, with no significant differences in the correlation strength between the sexes (Aβ42: males vs. females, z=0.90, P=0.370; p-Tau 181: males vs. females, z=0.60, P=0.550) (Figure 2E).

Figure 2 Associations between plasma Aβ42, p-Tau 181, and age. (A) Association between the level of plasma Aβ42 and age. (B) Association between the level of plasma p-Tau 181 and age. (C) Association between the level of plasma p-Tau 181 and Aβ42. (D) Relationship between the level of plasma Aβ42 (left)/p-Tau 181 (right) and age in the mild and severe WMH groups. (E) Relationship between the level of plasma Aβ42 (left)/p-Tau 181 (right) and age in the male and female participants. Aβ42, amyloid-beta 42; p-Tau 181, phosphorylated tau 181; WMH, white matter hyperintensity.

Associations between plasma AD pathological biomarkers and clinical conditions in different WMH severity groups

A positive correlation was found between the plasma p-Tau 181 concentrations and Fazekas scores. Conversely, no significant correlation was found between the Aβ42 levels and severity of WMH (Figure 3). After dividing the participants into mild and severe WMH groups based on their Fazekas scores, no significant correlations were found between the plasma AD pathological markers and sleep disorders, hypertension, type 2 diabetes, or prior LI in either group (Table 2 and Figure S1). However, in the severe WMH group, age was a potential risk factor for hypertension and prior LI (Table 2).

Figure 3 Relationship between the level of plasma Aβ42/p-Tau 181 and WMH severity. Box plots (left panel) showing the distribution of Aβ42 (A) and p-Tau 181 (B) across different Fazekas scores. The central line and box boundaries represent the median and interquartile range of values. The scatter plot (right panel) displays the partial correlation between the residuals of Aβ42 (A)/p-Tau 181 (B) and residuals of the Fazekas score after regressing out the effects of sex and age. The solid black line indicates the fitted linear regression line, with the shaded area representing the 95% confidence interval. Each point represents an individual participant. Aβ42, amyloid-beta 42; p-Tau 181, phosphorylated tau 181; WMH, white matter hyperintensity.

Associations between plasma AD pathological biomarkers and CSVD types in different WMH severity groups

As detailed in Table 3, the plasma Aβ42 and p-Tau 181 levels were not significantly correlated with the CSVD imaging markers in the mild WMH group, including CA, RSSI, and PVSs. The only significant finding was an inverse correlation between higher plasma Aβ42 levels and the presence of LI specifically in the severe WMH group (Figure S2). Conversely, increasing age was consistently associated with a higher likelihood of CA and LI in both WMH severity groups.

Differences in brain structure between the mild and severe WMH groups

Subcortical nuclei volumes were compared between the mild and severe WMH groups (Table 4). The left caudate nucleus was significantly larger in the severe WMH group after FDR correction. The right caudate also showed a significant volume increase in the severe WMH group before correction. Conversely, both the left and right putamen volumes were significantly reduced in the severe WMH group at the uncorrected threshold. However, no significant differences were observed in cortical thickness using only the original 3D T₁W images (Table S1).

Partial correlation analyses between plasma biomarkers, age, sex, and subcortical volumes

Based on the original 3D T₁W images, a significant positive correlation was found between the plasma Aβ42 levels and the right caudate nucleus volume, which survived correction for multiple comparisons. Meanwhile, the plasma p-Tau 181 levels were also found to be positively correlated with the left caudate volume at the uncorrected threshold. Additionally, age was significantly and negatively correlated with the volumes of most subcortical structures examined, including the thalamus, pallidum, hippocampus, amygdala, and nucleus accumbens (Figure 4).

Figure 4 Correlation matrix showing the partial correlation coefficients. (A) Heatmap displaying correlation coefficients between the plasma biomarkers (Aβ42/p-Tau 181) and subcortical brain volumes (adjusted for sex, age, and total intracranial volume). (B) Heatmap displaying the correlation coefficients between age or sex and subcortical brain volumes, adjusted for total intracranial volume and the other variable (sex/age). The color intensity represents the correlation strength (blue: negative, red: positive). The brain regions are ordered by the correlation strength, ^#, P<0.05, ^##, P<0.01, ^###, P<0.001; FDR-corrected *, q<0.05, **, q<0.01, ***, q<0.001. Aβ42, amyloid-beta 42; FDR, false discovery rate; p-Tau 181, phosphorylated tau 181.

Differences in caudate nucleus volumes and CSVD types between mild and severe WMHs at different ages

The participants were categorized into three age groups: <60, 60–75 and >75 years, and the bilateral caudate nucleus volumes between the mild and severe WMHs in the different age groups were compared. In the participants with original 3D T₁WI data, the left caudate volume began to exhibit significant differences between the mild and severe WMH groups from 60 years old (Figure 5A). Notably, no significant interaction was found between age and WMH group status in terms of the left caudate volume (β=1.597, P=0.203). However, when the reconstructed 3D T₁WI were included to increase the sample size, this interaction reached statistical significance (β=1.288, P=0.047) (Figure S3). Additionally, the difference in the frequency of the CSVD imaging markers between the mild and severe WMH groups was most noticeable in the 60–75-year age group, while it was less noticeable in the older age group (>75 years) (Figure 5B).

Figure 5 Bilateral caudate nucleus volumes and types of CSVD in different age groups. Quantitative data showing comparisons of the right and left caudate nucleus volumes (adjusted for sex and total intracranial volume) (A) and the prevalence of CSVD (B) between different WMH severity groups across different age ranges. The statistics for CA, RSSI, LI, and PVS were derived from the Chi-square test, and the statistics for concentrations of plasma Aβ42 and p-Tau 181 were derived from the Mann-Whitney U test, *, P<0.05; **, P<0.01. Aβ42, amyloid-beta 42; CA, cerebral atrophy; CSVD, cerebral small vessel disease; LI, lacunar infarction; p-Tau 181, phosphorylated tau 181; PVS, perivascular space; RSSI, recent small subcortical infarction; WMH, white matter hyperintensity.

Mediation analysis

All the 3D T₁WI data were included in the mediation analysis to increase the statistical effect (Figure 6A). ComBat processing was performed before integrating the original 3D T₁WI and reconstructed 3D T₁WI data (Figure S4). A significant indirect effect of plasma p-Tau 181 on left caudate volume mediated by the Fazekas score was observed (Figure 6B). The sensitivity analysis showed moderate stability with bootstrap CIs supporting the mediation effect (Figure 6C). To further investigate the reliability of our results, a repeat mediation analysis was performed using the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. The association between plasma p-Tau 181 and right caudate volume was significantly mediated by WMH volume (Figure S5).

Figure 6 Mediation model analysis. (A) Flowchart illustrating the data processing pipeline for integrating 2D and 3D T₁WI data for mediation analysis. (B) Path model testing the mediating role of WMH on the relationship between plasma p-Tau 181 and the left caudate nucleus volume in all participants, controlling for age, sex, and total intracranial volume. β represents effect, *, P<0.05; ***, P<0.001. (C) Bootstrap sensitivity analysis of the mediation effect. Distribution shows 500 bootstrap replications of the indirect effect, and the mean effect estimate and 95% confidence interval are displayed. The red line represents the original mediation effect shown in (B). 2D, two-dimensional; 3D, three-dimensional; CI, confidence interval; DWI, diffusion-weighted imaging; FLAIR, fluid-attenuated inversion recovery; MR, magnetic resonance; p-Tau 181, phosphorylated tau 181; T₁WI, T1-weighted imaging; T₂WI, T2-weighted imaging; WMH, white matter hyperintensity.

Discussion

The present study found a novel association among AD pathological biomarkers in plasma, WMH, and brain structural alterations during the aging process. The participants were stratified into mild and severe WMH groups based on their Fazekas scores, and a significant increase in the left caudate nucleus volume was observed in the severe WMH group. Additionally, this study found that WMH served as a partial mediator between plasma p-Tau 181 and the left caudate volume. Our results established a link between the pathology of AD and WMH throughout the aging process, unraveling their pathological roles in age-related structural brain changes.

WMH is a common radiological phenotype in aging brains and is increasingly considered an indicator of poor brain health. Individuals with a higher WMH burden tend to be older and exhibit a higher prevalence of CA, RSSI, and LI, suggesting that as age increases, there is a corresponding increase in the risk of CSVD, which is consistent with previous research findings (19,25). The participants with severe WMHs also exhibited significantly elevated plasma p-Tau 181 levels, as well as an upward trend in Aβ42 levels. Both plasma Aβ42 and p-Tau 181 levels have been reported to be correlated with WMH scores, and to enhance the predictive power of vascular risk factors in detecting the presence of CSVD and other vascular pathologies (26,27). However, we found that the correlation between plasma Aβ42 concentration and age was gender-specific, with a higher correlation in males. Conversely, the correlation between p-Tau 181 and age was equally high in males and females, which suggests that age-related p-Tau 181 abnormalities in plasma are more universal than those in Aβ42 in aging. Further, we found that in the severe WMH group, a decrease in plasma Aβ42 was associated with the occurrence of LI. Research has reported that vascular risk factors may indirectly contribute to the occurrence of LI through an increased WMH burden (28), whereas reduced Aβ42 levels may further amplify this risk by exacerbating endothelial dysfunction or neuroinflammation (29). These two processes may exert synergistic effects through shared vascular amyloid pathology, such as cerebral amyloid angiopathy, or through blood-brain barrier disruption (30).

In the severe WMH group, there was a significant increase in the bilateral caudate volumes, even though the enlargement in the right caudate did not survive correction. Previous studies have reported a positive correlation between the WMH load and caudate volumes in elderly individuals (31,32), which was substantiated by our observations of increased caudate volumes in the severe WMH group. Another study exploring broader age spans reported a U-shaped trajectory for caudate volumes, characterized by a decline from early adulthood through the sixth decade of life, followed by subsequent enlargement (33). The enlargement of the caudate nucleus in advanced age is probably concomitant with the presence of WMH. However, it has been observed that in middle-aged, cognitively unimpaired individuals at risk of AD, a larger WMH lesion volume is linked to smaller volumes in the right caudate nucleus (34). This inconsistency with our findings may also be influenced by the inverted U-shaped developmental trajectory of the caudate nucleus, given that our study involved an older population.

Notably, the caudate nucleus volume was found to be positively correlated with the plasma AD pathological markers. Although caudate atrophy has been documented in AD (35), no direct statistical association between plasma p-Tau 181 and caudate volume has been reported in cognitively normal elderly individuals. Additionally, under uncorrected criteria, the participants with severe WMH exhibited significant reductions in the bilateral putamen volume, which is consistent with previous research (36).

With increasing age, the differences in the left caudate volumes between the severe and mild WMH groups gradually increased, becoming significant from the age range of 60 to 75 years. When the sample size was increased, it was found that the interaction between age and WMH affected changes in the caudate nucleus volume. Although previous research on CSVD reported a notable volumetric decline with increasing global WMH load, predominately in the bilateral parietal lobes, anterior insula, and caudate nuclei (37,38), findings from other studies have indicated that CSVD and other vessel-related diseases are associated with significant enlargement of the caudate nucleus during the aging process (39-41). In addition, enlargement has also been observed in other neurological disorders, such as Parkinson’s disease and multiple sclerosis (42-44). These results illustrate the susceptibility of the caudate nucleus in certain CNS diseases, particularly under the influence of prevalent CSVD (45,46). Indeed, a recent study showed that iron accumulation and gliosis in the brain, especially the caudate nucleus, are significantly associated with memory decline in the elderly (47). The enlargement of the caudate nucleus observed in this study is likely to be the result of gliosis. However, this enlargement in individuals with AD is thought to be a compensatory reaction to decreased hippocampal function (48,49). Therefore, further research is needed to fully understand the specific mechanisms and effects of caudate nucleus enlargement in severe WMHs. Moreover, the differences in CSVD radiological features, such as LI and CA, between mild and severe WMHs were less pronounced after the age of 75 years. This is because the prevalence increased significantly in the oldest age range in the mild WMH group, reducing the difference between the two groups (Figure S6).

The mediation analysis revealed that WMH mediated the relationship between plasma p-Tau 181 and the caudate nucleus volume, a finding that was reproduced and validated in an independent cohort of cognitively normal older adults from the ADNI database in our study. The effect was observed in the right caudate nucleus in the ADNI cohort; however, this may be due to the use of the total WMH volume rather than the Fazekas score as the mediating variable. Nevertheless, WMH appears to play a role in the relationship between plasma p-Tau 181 and the caudate nucleus. Specifically, plasma p-Tau 181 facilitates an increase in the caudate nucleus volume, partially through WMH. Meso-scale discovery and single-molecule array (Simoa) platform measurements showed that a high plasma concentration of p-Tau 181 is associated with greater WMH volume in cognitively unimpaired elderly individuals (13). An association has also been found between p-Tau 181 in CSF and WMH in non-demented older participants, which may be influenced by apolipoprotein E (APOE) ε4 carrier status and cerebrovascular risk factors (50). These findings highlight that elevated levels of p-Tau 181 are linked to increased WMH severity. Given the correlation between WMH and the caudate nucleus volume mentioned above, it is highly likely that WMH plays a significant mediating role in the relationship between tau pathology and caudate nucleus in aging.

Notably, we conducted a segmented regression analysis and found that the threshold at which plasma p-Tau 181 influenced the left caudate volume was 16.5 pg/mL (data not shown). This level is slightly higher than previously reported concentrations in cognitively normal individuals. To validate this finding, we examined data from cognitively normal adults in the ADNI database, where the mean plasma p-Tau 181 concentration was 16.78 pg/mL, which suggests that even cognitively normal older adults may have biomarker levels that fall on the AD continuum. Thus, the influence of plasma p-Tau 181 on the caudate nucleus may, to some extent, involve overlapping mechanisms with the pathological processes of AD. Since p-Tau 181 in plasma is positively linked to amyloid-PET, the relationship between AD-specific biomarkers and WMH may be explained by cerebral amyloid angiopathy (27). AD patients with severe cerebral amyloid angiopathy have been shown to exhibit increased brain iron levels, suggesting that microvascular degeneration may play a key role in iron deposition in AD (45). Meanwhile, the caudate nucleus is particularly susceptible to iron and amyloid deposition. Similarly, the abnormally increased function of the caudate nucleus in AD patients further suggests its susceptibility (51,52).

Our results showed that tau pathology promoted microvascular degeneration and caused structural abnormalities in the left caudate nucleus in the majority of the elderly population. Conversely, plasma Aβ42 was not found to have a statistically significant effect in the mediation analysis. Similarly, research has shown that plasma Aβ42 cannot be used to predict amyloid-PET (53). Thus, amyloid in the brain may be the pathogenic mechanism underlying the interaction between plasma p-Tau 181, WMH, and the left caudate nucleus. Moreover, the enlargement of the left caudate nucleus was more pronounced with advancing age, which may be because amyloid deposition and microvascular degeneration become more severe as age increases. Future studies need to be conducted with elderly participants to assess this hypothesis.

This study had several limitations that should be noted. First, the data collected originated from routine clinical diagnostics, and detailed cognitive assessment data were lacking, which prevented us from observing the effect of the enlargement of the left caudate nucleus on cognition. Second, since some T₂ FLAIR imaging data were 2D diagnostic images, we were unable to calculate the quantitative volume of WMH and could only measure the extent of white matter damage using the Fazekas scale. Third, while more sensitive detection methods such as SIMOA and mass spectrometry are available, the approach employed in this study for measuring plasma biomarkers was the conventional ELISA. However, the clinical performances of the ELISA and SIMOA platforms were comparable for plasma Aβ42 and t-Tau (54).

Conclusions

In the present study, we simultaneously investigated WMH, plasma AD-related biomarkers, and brain structural alterations, focusing on their potential associations in cognitively normal elderly individuals. We found that elderly individuals with a high burden of WMH exhibited a significant increase in the caudate nucleus volume. Additionally, higher plasma p-Tau 181 levels were associated with caudate nucleus enlargement, partially mediated by an increased WMH burden. Future studies should explore whether this relationship is due to amyloid deposition in the caudate nucleus. These results suggest that WMH may serve as a mediating pathway linking cerebrovascular pathology to AD pathology in the aging process of the brain, highlighting the importance of early vascular risk management.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-376/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-376/dss

Funding: This work was supported by the National Science and Technology Innovation 2030 -- Major Program of “Brain Science and Brain-Like Research (No. 2022ZD0211800), the National Natural Science Foundation of China (Nos. t2330059, 82271965, and 82302172), General Project Supported by Medical Science and Technology development Foundation, Nanjing Department of Health (Nos. YKK22083 and YKK23103) and the Jiangsu Funding Program for Excellent Postdoctoral Talent (No. 2022ZB694) to B.Z.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-376/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Ethics Committee of Nanjing Drum Tower Hospital (IRB review approval No. 2022-165-01). Informed consent was obtained from all participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Indahlastari A, Albizu A, Boutzoukas EM, O'Shea A, Woods AJ. White matter hyperintensities affect transcranial electrical stimulation in the aging brain. Brain Stimul 2021;14:69-73. [Crossref] [PubMed]
2. Ottavi TP, Pepper E, Bateman G, Fiorentino M, Brodtmann A. Consensus statement for the management of incidentally found brain white matter hyperintensities in general medical practice. Med J Aust 2023;219:278-84. [Crossref] [PubMed]
3. Cha WJ, Yi D, Ahn H, Byun MS, Chang YY, Choi JM, Kim K, Choi H, Jung G, Kang KM, Sohn CH, Lee YS, Kim YK, Lee DY. Association between brain amyloid deposition and longitudinal changes of white matter hyperintensities. Alzheimers Res Ther 2024;16:50. [Crossref] [PubMed]
4. Garnier-Crussard A, Cotton F, Krolak-Salmon P, Chételat G. White matter hyperintensities in Alzheimer's disease: Beyond vascular contribution. Alzheimers Dement 2023;19:3738-48. [Crossref] [PubMed]
5. Blinkouskaya Y, Caçoilo A, Gollamudi T, Jalalian S, Weickenmeier J. Brain aging mechanisms with mechanical manifestations. Mech Ageing Dev 2021;200:111575. [Crossref] [PubMed]
6. Roe JM, Vidal-Piñeiro D, Sørensen Ø, Grydeland H, Leonardsen EH, Iakunchykova O, et al. Brain change trajectories in healthy adults correlate with Alzheimer's related genetic variation and memory decline across life. Nat Commun 2024;15:10651. [Crossref] [PubMed]
7. Suzuki H, Venkataraman AV, Bai W, Guitton F, Guo Y, Dehghan A, Matthews PMAlzheimer’s Disease Neuroimaging Initiative. Associations of Regional Brain Structural Differences With Aging, Modifiable Risk Factors for Dementia, and Cognitive Performance. JAMA Netw Open 2019;2:e1917257. [Crossref] [PubMed]
8. Schulz M, Mayer C, Schlemm E, Frey BM, Malherbe C, Petersen M, Gallinat J, Kühn S, Fiehler J, Hanning U, Twerenbold R, Gerloff C, Cheng B, Thomalla G. Association of Age and Structural Brain Changes With Functional Connectivity and Executive Function in a Middle-Aged to Older Population-Based Cohort. Front Aging Neurosci 2022;14:782738. [Crossref] [PubMed]
9. Mielke MM, Dage JL, Frank RD, Algeciras-Schimnich A, Knopman DS, Lowe VJ, Bu G, Vemuri P, Graff-Radford J, Jack CR Jr, Petersen RC. Performance of plasma phosphorylated tau 181 and 217 in the community. Nat Med 2022;28:1398-405. [Crossref] [PubMed]
10. Wilson EN, Young CB, Ramos Benitez J, Swarovski MS, Feinstein I, Vandijck M, et al. Performance of a fully-automated Lumipulse plasma phospho-tau181 assay for Alzheimer's disease. Alzheimers Res Ther 2022;14:172. [Crossref] [PubMed]
11. Bolton CJ, Steinbach M, Khan OA, Liu D, O'Malley J, Dumitrescu L, Peterson A, Jefferson AL, Hohman TJ, Zetterberg H, Gifford KAAlzheimer’s Disease Neuroimaging Initiative. Clinical and demographic factors modify the association between plasma phosphorylated tau-181 and cognition. Alzheimers Dement (Amst) 2024;16:e70047. [Crossref] [PubMed]
12. Yu L, Boyle PA, Janelidze S, Petyuk VA, Wang T, Bennett DA, Hansson O, Schneider JA. Plasma p-tau181 and p-tau217 in discriminating PART, AD and other key neuropathologies in older adults. Acta Neuropathol 2023;146:1-11. [Crossref] [PubMed]
13. Mielke MM, Frank RD, Dage JL, Jeromin A, Ashton NJ, Blennow K, Karikari TK, Vanmechelen E, Zetterberg H, Algeciras-Schimnich A, Knopman DS, Lowe V, Bu G, Vemuri P, Graff-Radford J, Jack CR Jr, Petersen RC. Comparison of Plasma Phosphorylated Tau Species With Amyloid and Tau Positron Emission Tomography, Neurodegeneration, Vascular Pathology, and Cognitive Outcomes. JAMA Neurol 2021;78:1108-17. [Crossref] [PubMed]
14. Sun Y, Hu HY, Hu H, Huang LY, Tan L, Yu JTAlzheimer’s Disease Neuroimaging Initiative. Cerebral Small Vessel Disease Burden Predicts Neurodegeneration and Clinical Progression in Prodromal Alzheimer's Disease. J Alzheimers Dis 2023;93:283-94. [Crossref] [PubMed]
15. Laing KK, Simoes S, Baena-Caldas GP, Lao PJ, Kothiya M, Igwe KC, Chesebro AG, Houck AL, Pedraza L, Hernández AI, Li J, Zimmerman ME, Luchsinger JA, Barone FC, Moreno H, Brickman AMAlzheimer’s Disease Neuroimaging Initiative. Cerebrovascular disease promotes tau pathology in Alzheimer's disease. Brain Commun 2020;2:fcaa132. [Crossref] [PubMed]
16. Tissot C, L, Benedet A, Therriault J, Pascoal TA, Lussier FZ, Saha-Chaudhuri P, et al. Plasma pTau181 predicts cortical brain atrophy in aging and Alzheimer's disease. Alzheimers Res Ther 2021;13:69. [Crossref] [PubMed]
17. Rizvi B, Narkhede A, Last BS, Budge M, Tosto G, Manly JJ, Schupf N, Mayeux R, Brickman AM. The effect of white matter hyperintensities on cognition is mediated by cortical atrophy. Neurobiol Aging 2018;64:25-32. [Crossref] [PubMed]
18. Al-Janabi OM, Bauer CE, Goldstein LB, Murphy RR, Bahrani AA, Smith CD, Wilcock DM, Gold BT, Jicha GA. White Matter Hyperintensity Regression: Comparison of Brain Atrophy and Cognitive Profiles with Progression and Stable Groups. Brain Sci 2019;9:170. [Crossref] [PubMed]
19. Chojdak-Łukasiewicz J, Dziadkowiak E, Zimny A, Paradowski B. Cerebral small vessel disease: A review. Adv Clin Exp Med 2021;30:349-56. [Crossref] [PubMed]
20. Lancaster JL, Woldorff MG, Parsons LM, Liotti M, Freitas CS, Rainey L, Kochunov PV, Nickerson D, Mikiten SA, Fox PT. Automated Talairach atlas labels for functional brain mapping. Hum Brain Mapp 2000;10:120-31. [Crossref] [PubMed]
21. Destrieux C, Fischl B, Dale A, Halgren E. Automatic parcellation of human cortical gyri and sulci using standard anatomical nomenclature. Neuroimage 2010;53:1-15. [Crossref] [PubMed]
22. Liu G, Cao Z, Xu Q, Zhang Q, Yang F, Xie X, Hao J, Shi Y, Bernhardt B C, He Y, Shi F, Lu G, Zhang Z. Recycling diagnostic MRI for empowering brain morphometric research - Critical & practical assessment on learning-based image super-resolution. Neuroimage 2021;245:118687. [Crossref] [PubMed]
23. Fortin JP, Cullen N, Sheline YI, Taylor WD, Aselcioglu I, Cook PA, Adams P, Cooper C, Fava M, McGrath PJ, McInnis M, Phillips ML, Trivedi MH, Weissman MM, Shinohara RT. Harmonization of cortical thickness measurements across scanners and sites. Neuroimage 2018;167:104-20. [Crossref] [PubMed]
24. Orlhac F, Lecler A, Savatovski J, Goya-Outi J, Nioche C, Charbonneau F, Ayache N, Frouin F, Duron L, Buvat I. How can we combat multicenter variability in MR radiomics? Validation of a correction procedure. Eur Radiol 2021;31:2272-80. [Crossref] [PubMed]
25. Li T, Huang Y, Cai W, Chen X, Men X, Lu T, Wu A, Lu Z. Age-related cerebral small vessel disease and inflammaging. Cell Death Dis 2020;11:932. [Crossref] [PubMed]
26. Qu W, Zhang L, Liang X, Yu Z, Huang H, Zhao J, Guo Y, Zhou X, Xu S, Luo H, Luo X. Elevated Plasma Oligomeric Amyloid β-42 Is Associated with Cognitive Impairments in Cerebral Small Vessel Disease. Biosensors (Basel) 2023;13:110. [Crossref] [PubMed]
27. Twait EL, Gerritsen L, Moonen JEF, Verberk IMW, Teunissen CE, Visser PJ, van der Flier WM, Geerlings MIUCC SMART Study Group, the NCDC Consortium. Plasma Markers of Alzheimer's Disease Pathology, Neuronal Injury, and Astrocytic Activation and MRI Load of Vascular Pathology and Neurodegeneration: The SMART-MR Study. J Am Heart Assoc 2024;13:e032134. [Crossref] [PubMed]
28. Hussainali RF, Adank MC, Lamballais S, Vernooij MW, Ikram MA, Steegers EAP, Miller EC, Schalekamp-Timmermans S. Hypertension in Pregnancy Linked to Cerebral Small Vessel Disease 15 Years Later. Hypertension 2025;82:1326-35. [Crossref] [PubMed]
29. Howe MD, Caruso MR, Manoochehri M, Kunicki ZJ, Emrani S, Rudolph JL, Huey ED, Salloway SP, Oh HAlzheimer’s Disease Neuroimaging Initiative. Utility of cerebrovascular imaging biomarkers to detect cerebral amyloidosis. Alzheimers Dement 2024;20:7220-31. [Crossref] [PubMed]
30. Charidimou A, Baron JC. White Matter Hyperintensity Multispot Pattern Lesions and Cerebrovascular Amyloid Burden in Cerebral Amyloid Angiopathy. Stroke 2025;56:2767-71. [Crossref] [PubMed]
31. Song H, Bharadwaj PK, Grilli MD, Raichlen DA, Habeck CG, Huentelman MJ, Hishaw GA, Trouard TP, Alexander GE. Subcortical brain volumetric differences related to white matter lesion volume and cognition in healthy aging. NPJ Aging 2025;11:44. [Crossref] [PubMed]
32. Macfarlane MD, Looi JC, Walterfang M, Spulber G, Velakoulis D, Styner M, Crisby M, Orndahl E, Erkinjuntti T, Waldemar G, Garde E, Hennerici MG, Bäzner H, Blahak C, Wallin A, Wahlund LOLADIS Study Group. Shape abnormalities of the caudate nucleus correlate with poorer gait and balance: results from a subset of the LADIS study. Am J Geriatr Psychiatry 2015;23:59-71.e1. [Crossref] [PubMed]
33. Potvin O, Mouiha A, Dieumegarde L, Duchesne SAlzheimer's Disease Neuroimaging Initiative. Normative data for subcortical regional volumes over the lifetime of the adult human brain. Neuroimage 2016;137:9-20. [Crossref] [PubMed]
34. Brugulat-Serrat A, Salvadó G, Operto G, Cacciaglia R, Sudre CH, Grau-Rivera O, Suárez-Calvet M, Falcon C, Sánchez-Benavides G, Gramunt N, Minguillon C, Fauria K, Barkhof F, Molinuevo JL, Gispert JD. ALFA Study. White matter hyperintensities mediate gray matter volume and processing speed relationship in cognitively unimpaired participants. Hum Brain Mapp 2020;41:1309-22. [Crossref] [PubMed]
35. Wurst Z, Birčák Kuchtová B, Křemen J, Lahutsina A, Ibrahim I, Tintěra J, Bartoš A, Brabec M, Rai T, Zach P, Musil V, Olympiou N, Mrzílková J. Basal Ganglia Compensatory White Matter Changes on DTI in Alzheimer's Disease. Cells 2023;12:1220. [Crossref] [PubMed]
36. Sleight E, Stringer MS, Clancy U, Arteaga C, Jaime Garcia D, Hewins W, et al. Cerebrovascular Reactivity in Patients With Small Vessel Disease: A Cross-Sectional Study. Stroke 2023;54:2776-84. [Crossref] [PubMed]
37. Lambert C, Sam Narean J, Benjamin P, Zeestraten E, Barrick TR, Markus HS. Characterising the grey matter correlates of leukoaraiosis in cerebral small vessel disease. Neuroimage Clin 2015;9:194-205. [Crossref] [PubMed]
38. Camarda C, Torelli P, Pipia C, Battaglini I, Azzarello D, Rosano R, Ventimiglia GD, Sottile G, Cilluffo G, Camarda R. Association Between Atrophy of the Caudate Nuclei, Global Brain Atrophy, Cerebral Small Vessel Disease and Mild Parkinsonian Signs in Neurologically and Cognitively Healthy Subjects Aged 45-84 Years: A Crosssectional Study. Curr Alzheimer Res 2018;15:1013-26. [Crossref] [PubMed]
39. Liu S, Hou B, You H, Zhang Y, Zhu Y, Ma C, Zuo Z, Feng F. The Association Between Perivascular Spaces and Cerebral Blood Flow, Brain Volume, and Cardiovascular Risk. Front Aging Neurosci 2021;13:599724. [Crossref] [PubMed]
40. Zhao J, Kong Q, Zhou X, Zhang Y, Yu Z, Qu W, Huang H, Luo X. Differences in Gray Matter Volume in Cerebral Small Vessel Disease Patients with and without Sleep Disturbance. Brain Sci 2023;13:294. [Crossref] [PubMed]
41. Husøy AK, Pintzka C, Eikenes L, Håberg AK, Hagen K, Linde M, Stovner LJ. Volume and shape of subcortical grey matter structures related to headache: A cross-sectional population-based imaging study in the Nord-Trøndelag Health Study. Cephalalgia 2019;39:173-84. [Crossref] [PubMed]
42. Kaut O, Mielacher C, Hurlemann R, Wüllner U. Resting-state fMRI reveals increased functional connectivity in the cerebellum but decreased functional connectivity of the caudate nucleus in Parkinson's disease. Neurol Res 2020;42:62-7. [Crossref] [PubMed]
43. Kim K, Bohnen NI, Müller MLTM, Lustig C. Compensatory dopaminergic-cholinergic interactions in conflict processing: Evidence from patients with Parkinson's disease. Neuroimage 2019;190:94-106. [Crossref] [PubMed]
44. Bardel B, Chalah MA, Bensais-Rueda R, Créange A, Lefaucheur JP, Ayache SS. Event-related desynchronization and synchronization in multiple sclerosis. Mult Scler Relat Disord 2024;86:105601. [Crossref] [PubMed]
45. Sun C, Wu Y, Ling C, Xie Z, Kong Q, Fang X, An J, Sun Y, Zhang W, Yang Q, Wang Z, Zhang Z, Yuan Y. Deep Gray Matter Iron Deposition and Its Relationship to Clinical Features in Cerebral Autosomal Dominant Arteriopathy With Subcortical Infarcts and Leukoencephalopathy Patients: A 7.0-T Magnetic Resonance Imaging Study. Stroke 2020;51:1750-7. [Crossref] [PubMed]
46. Chai C, Wang H, Liu S, Chu ZQ, Li J, Qian T, Haacke EM, Xia S, Shen W. Increased iron deposition of deep cerebral gray matter structures in hemodialysis patients: A longitudinal study using quantitative susceptibility mapping. J Magn Reson Imaging 2019;49:786-99. [Crossref] [PubMed]
47. Venkatesh A, Daugherty AM, Bennett IJ. Neuroimaging measures of iron and gliosis explain memory performance in aging. Hum Brain Mapp 2021;42:5761-70. [Crossref] [PubMed]
48. Persson K, Bohbot VD, Bogdanovic N, Selbaek G, Braekhus A, Engedal K. Finding of increased caudate nucleus in patients with Alzheimer's disease. Acta Neurol Scand 2018;137:224-32. [Crossref] [PubMed]
49. Sodums DJ, Bohbot VD. Negative correlation between grey matter in the hippocampus and caudate nucleus in healthy aging. Hippocampus 2020;30:892-908. [Crossref] [PubMed]
50. Newton P, Tchounguen J, Pettigrew C, Lim C, Lin Z, Lu H, Moghekar A, Albert M, Soldan ABIOCARD Research Team. Regional White Matter Hyperintensities and Alzheimer's Disease Biomarkers Among Older Adults with Normal Cognition and Mild Cognitive Impairment. J Alzheimers Dis 2023;92:323-39. [Crossref] [PubMed]
51. Li J, Jin D, Li A, Liu B, Song C, Wang P, et al. ASAF: altered spontaneous activity fingerprinting in Alzheimer's disease based on multisite fMRI. Sci Bull (Beijing) 2019;64:998-1010. [Crossref] [PubMed]
52. Chen B, Wang Q, Zhong X, Mai N, Zhang M, Zhou H, Haehner A, Chen X, Wu Z, Auber LA, Rao D, Liu W, Zheng J, Lin L, Li N, Chen S, Chen B, Hummel T, Ning Y. Structural and Functional Abnormalities of Olfactory-Related Regions in Subjective Cognitive Decline, Mild Cognitive Impairment, and Alzheimer's Disease. Int J Neuropsychopharmacol 2022;25:361-74. [Crossref] [PubMed]
53. Fowler CJ, Stoops E, Rainey-Smith SR, Vanmechelen E, Vanbrabant J, Dewit N, Mauroo K, Maruff P, Rowe CC, Fripp J, Li QX, Bourgeat P, Collins SJ, Martins RN, Masters CL, Doecke JD. Plasma p-tau181/Aβ(1-42) ratio predicts Aβ-PET status and correlates with CSF-p-tau181/Aβ(1-42) and future cognitive decline. Alzheimers Dement (Amst) 2022;14:e12375. [Crossref] [PubMed]
54. De Meyer S, Schaeverbeke JM, Verberk IMW, Gille B, De Schaepdryver M, Luckett ES, Gabel S, Bruffaerts R, Mauroo K, Thijssen EH, Stoops E, Vanderstichele HM, Teunissen CE, Vandenberghe R, Poesen K. Comparison of ELISA- and SIMOA-based quantification of plasma Aβ ratios for early detection of cerebral amyloidosis. Alzheimers Res Ther 2020;12:162. [Crossref] [PubMed]