The glymphatic system is a recently discovered waste drainage system in the brain that involves movement of cerebrospinal fluid (CSF) through the perivascular and interstitial spaces (Taoka and Naganawa, 2020). The perivascular space around the arteries allows CSF to enter the interstitial spaces through water channels controlled by aquaporin 4. CSF entering the interstitial space washes away waste proteins within the tissues, such as amyloid-β (Aβ) and metabolites, and facilitates the distribution of glucose, lipids, amino acids, and neuromodulators (Iliff et al., 2012, 2013; Jessen et al., 2015). Therefore, the glymphatic system is considered to play an important role in many neurological diseases including Alzheimer’s disease (AD), normal pressure hydrocephalus, sleep disorder, and vascular disease (Taoka et al., 2017; Jiang, 2019).

Evaluation of the glymphatic system has been performed with fluorescent tracers in animal experiments (Iliff et al., 2012, 2013; Jessen et al., 2015). As radiotracer studies are limited in the human population, many previous studies utilized magnetic resonance imaging (MRI) techniques to measure and visualize the CNS fluid flow compatible with the glymphatic system, primarily with gadolinium-based contrast agents (Lee M. et al., 2022). However, complex MRI acquisition including invasive intrathecal or intravenous contrast agent injection still limits widespread clinical application. Recently, a non-invasive method called “Diffusion Tensor Image-Analysis aLong the Perivascular Space” (DTI-ALPS) has been introduced for glymphatic assessment to overcome the drawbacks of contrast enhanced MRI (Taoka et al., 2017). With this method, the motion of water molecules in the direction of perivascular spaces can be evaluated by measuring diffusivity using the diffusion tensor method (Taoka et al., 2017). This DTI-based method has effectively shown altered glymphatic function in many neurological diseases (Bae et al., 2021; Chen et al., 2021; Steward et al., 2021; Kamagata et al., 2022; Lee D. et al., 2022; Ruan et al., 2022). Many previous studies have assessed, measured, and visualized normal CNS fluid flow compatible with the glymphatic hypothesis in human participants (Lee M. et al., 2022). However, few studies have applied DTI-ALPS for evaluation of glymphatic activity in non-diseased population.

Subjective cognitive decline (SCD) is the self-reported experience of impaired memory but with normal cognitive performance on an objective neuropsychological test (Jessen et al., 2014). Although the results of some studies are inconsistent, biomarker studies of SCD including neuroimaging generally suggest SCD as an early prodromal stage of AD that precedes mild cognitive impairment (MCI). This is based on the similar underlying pathological changes in SCD to those in MCI and dementia (Jessen et al., 2014; Parker et al., 2022). Based on previous studies that revealed an impaired glymphatic system in MCI and AD, we hypothesized that the glymphatic system would be decreased in subjects with SCD with normal cognition, and that DTI-ALPS may capture this impairment.

The purpose of the present study was to assess the glymphatic system in cognitively normal older adults with or without SCD using DTI-ALPS, which is an indirect measurement of the glymphatic system. For further understanding of the glymphatic system, we correlated DTI-ALPS results with amyloid PET and aimed to identify relevant clinical and imaging features showing significant associations with DTI-ALPS.

This study aimed to evaluate the glymphatic system in cognitively normal older adults using non-invasive DTI by correlating it with clinical variables and amyloid PET. There were no significant differences in DTI-ALPS indices between adults with or without SCD. The greater amyloid burden in the paracentral cortex and older age were significantly associated with a lower ALPS index, indicating impaired glymphatic clearance in cognitively normal older adults.

Intrathecal or intravenous contrast injection for evaluation of the glymphatic system is invasive and poses risks for gadolinium deposition in the brain (Öner et al., 2017). Therefore, the DTI-ALPS method was recently developed utilizing non-invasive DTI without need for a contrast agent for assessment of the glymphatic system (Taoka et al., 2017). As DTI can be rapidly obtained while allowing multiple image acquisitions in a single subject, it enables monitoring of the status of the glymphatic activity over time. DTI-ALPS has been applied in neurodegenerative diseases including AD (Steward et al., 2021) and Parkinson’s disease (Chen et al., 2021), and the greater extent of impaired glymphatic clearance was noted in patients with more severe degrees of cognitive decline. Glymphatic function assessed with DTI-ALPS was also impaired in patients with normal pressure hydrocephalus (Bae et al., 2021), Parkinson’s disease (Ruan et al., 2022), and isolated REM sleep behavior disorder (Lee D. et al., 2022) compared to normal controls. Therefore, DTI-ALPS may be a feasible, non-invasive tool for the evaluation of the glymphatic system. In this study, we evaluated whether cognitively normal older adults who complained of SCD showed a significantly impaired glymphatic system compared to asymptomatic adults. However, there were no significant differences between diffusivities or ALPS index between adults with and without SCD. As SCD is self-experienced decline in cognitive function without evidence of objective cognitive impairment, the degree of impairment in glymphatic clearances may be too subtle to be captured in the ALPS index. Therefore, DTI-ALPS may not be useful as an indirect measurement of glymphatic activity in cognitively unimpaired adults with SCD. Future studies with larger number of adults with SCD may be needed for evaluation of impairment in glymphatic clearances.

This study investigated the correlation between amyloid deposition and the ALPS index, and found that the amyloid deposition, particularly in the paracentral cortex, was significantly associated with a lower ALPS index. ALPS index, which is the measurement of diffusivity along the direction of perivascular spaces, has been regarded as an indirect measurement of glymphatic activity as the main stream of glymphatic system involves movement of CSF along the perivascular spaces. Meanwhile, the key concept of glymphatic system is that after subarachnoid CSF enters the brain through periarterial spaces and mixes with interstitial fluid, it drains through perivenous spaces, meningeal lymphatic vessels (MLVs), or perineural pathways and finally reach at the deep cervical lymph nodes (Zhang et al., 1990; Iliff et al., 2013; Jessen et al., 2015; Taoka et al., 2017; Naganawa et al., 2020). As all these key elements in the glymphatic system correlate and regulate each other significantly, alteration in one component either due to aging or neurodegenerative diseases can affect the remainder profoundly. MLVs are one of the main routes for CSF efflux in the glymphatic system (Louveau et al., 2017; Jiang et al., 2022), and is well-known that these vessels experience functional decline with aging (Da Mesquita et al., 2021). Several previous studies with MRI reported their dorsal distribution along the venous sinuses that they are around almost all dural venous-parasagittal structures (Kuo et al., 2018; Albayram et al., 2022). Several recent studies with mice have suggested that either deterioration or ablation of MLVs results in increased amyloid deposition, specifically along the dura adjacent to the superior sagittal sinus (Da Mesquita et al., 2018; Wang et al., 2019). One study also reported significant amyloid deposition in the dura mater adjacent to the superior sagittal sinus in AD patients compared to normal controls (Da Mesquita et al., 2018). The authors concluded that prominent meningeal amyloid deposition occurs in AD patients and in mouse models of AD after MLV ablation. Even though our study subjects did not include patients with dementia, it can be inferred from previous studies that damage of function in MLVs, either due to aging or neurodegenerative diseases, can result in subsequent amyloid deposition in the adjacent structures along the superior sagittal sinus. As each compartment of glymphatic system is closely correlated with each other (Da Mesquita et al., 2018), impaired drainage through MLVs may affect overall glymphatic activity at the level of perivascular spaces, which can be captured by decreased ALPS index. Or, conversely, impaired glymphatic activity at the level of perivascular spaces may in turn affects the function of MLVs, which may attribute to the greater amyloid burden in the parasagittal regions including paracentral cortex. The reason why the specific paracentral cortex, rather than other regions along the superior sagittal sinus, showed significant association with ALPS index is yet elucidated. The reason for the lack of association of amyloid deposition in the posterior cingulate cortex or medial orbitofrontal cortex, where frequent amyloid deposits are encountered in AD, with DTI-ALPS is also not clear. Future studies with larger subjects including patients with objective cognitive decline may give insight regarding the association between the indirect measurement of glymphatic activity, i.e., ALPS index and amyloid deposition in particular brain regions. We believe that our study has its novelty as this is the first study that correlated the ALPS with amyloid PET findings in cognitively normal older adults.

This study found that age was the only clinical variable that was negatively correlated with the ALPS index. Similar results were observed in several previous studies that evaluated the impact of clinical demographics and vascular risk factors on the ALPS index in healthy subjects (Zhou et al., 2020; Zhang et al., 2021). During aging, the number of polarized aquaporin-4 channels on the end feet of astrocyte cells and CSF production decrease along with arterial pulsatility, attributed to the drop of glymphatic function (Jessen et al., 2015). In one recent publication, the ALPS index was also lower in males and patients with hypertension (Zhang et al., 2021). However, the ALPS index was not affected by these variables in the present study. These inconsistent findings may be attributed to different demographic characteristics, as the study subjects were older than those of the previous study (Zhang et al., 2021) (mean age 73.4 vs. 60.8 years) and had a higher proportion of adults with hypertension (45.6 vs. 39.4%). A larger number of normal older adults recruited from heterogeneous backgrounds is needed to evaluate the impact of clinical variables on the ALPS index.

A previous study, which first introduced the DTI-ALPS method for the evaluation of glymphatic activity, reported a significant positive correlation between the ALPS index and MMSE scores, indicating lower water diffusivity along the perivascular space in relation to AD severity (Taoka et al., 2017). Another study also reported significant correlations between DTI-ALPS and MMSE after adjustment of clinically relevant variables (Steward et al., 2021). However, the current study did not identify any significant correlation between MMSE scores and the ALPS index. As it included only adults with objectively normal cognition, excluding patients with documented cognitive decline, the differences in MMSE scores among subjects might be subtle. Most enrolled subjects from previous studies were patients with variable degrees of cognitive decline, demonstrating significantly lower MMSE scores compared to those in the present study.

According to the specific cut-off of SUVR from amyloid PET, 112 and 11 study subjects were classified into amyloid negative and positive group, respectively. As we have only enrolled elderly adults without objective cognitive decline, the number of subjects who had amyloid positivity was inevitably small. We observed that there was no statistically significant difference in ALPS index between the two groups, however, future studies with larger number of study subjects should validate our study results. In addition, further studies including patients with documented cognitive decline may give insight with regard to association between ALPS index and amyloid positivity.

Our study had several limitations. First, we only enrolled cognitively normal older adults with and without SCD and excluded patients with objective cognitive decline. The relationship between cognitive scores and the ALPS index may be reliably assessed in future studies including larger numbers of both cognitively normal and cognitively impaired adults. Second, as we performed cross-sectional analysis, it was not possible to examine the clinical significance of amyloid accumulation in cognitively normal individuals. In fact, there have been inconsistencies regarding the significance of amyloid deposition on future cognitive decline in the cognitively normal populations (Chételat et al., 2013; Guo et al., 2020). In addition, there is no established cutoff for SUVR from each cortical region that determines the significance of amyloid deposition. The SUVRs were correlated with the ALPS index from all cortical regions in a form of continuous variables. In this way, a specific cortical region was identified that may have an impact on the glymphatic system. Third, the DTI-ALPS was only used for assessment of glymphatic activity, and contrast-enhanced imaging methods were not performed in this study. Contrast-enhanced MRI may be superior to non-contrast MRI for glymphatic imaging, as it provides objective measurement of the degree of glymphatic flow changes (Lee M. et al., 2022). However, acquisition of contrast-enhanced sequences in normal individuals was not ethical. Despite these limitations, the researchers believe that DTI-ALPS is a feasible method for in vivo assessment of the glymphatic system, as demonstrated in many previous studies. Fourth, we were not able to include other behavioral risk factors for dementia (e.g., exercise, sleep disturbances, or social activity) in our study. The associations between ALPS index and other dementia risk factors should be studied for evaluation of clinical relevance of DTI-ALPS.

In conclusion, DTI-ALPS may not be useful for the evaluation of the glymphatic system in subjects with SCD. Older age was significantly associated with lower ALPS index. Greater amyloid deposition in the paracentral cortex was significantly associated with lower glymphatic activity in cognitively normal older adults. These results should be validated in future studies on the relationships between ALPS index, perivenous space, and the meningeal lymphatic drainage system.

The datasets presented in this article are not readily available because unable to seek consent from research subjects for external disclosure of original data. Requests to access the datasets should be directed to WJK, woojungkim@yuhs.ac.

The Institutional Review Board of Yongin Severance Hospital approved this study (9-2020-0080). The patients/participants provided their written informed consent to participate in this study.

CJP performed the conceptualization, formal analysis, and visualization of data and wrote and revised the manuscript. S-YK performed the preprocessing for DTI-ALPS analysis and calculated the ALPS index. JHK supported the formal analysis and visualization of data. N-HS contributed to the process of statistical complementation in the major revision process. JP contributed to the grouping of subjects and fully analyzed the PET imaging data and major revision. JYP supported the research progress through additional funding and supervised the whole project. YHJ and HJK contributed to the acquisition of PET imaging data. WJK performed the major roles in conceptualization, supervision, writing the manuscript, data curation, funding acquisition, and project administration. All authors had approved the submitted version of the manuscript.