Glymphatic alteration in NAFLD patient: a preliminary magnetic resonance imaging study based on DTI-ALPS

Abstract

Objective

This study aims to investigate alterations in glymphatic system (GS) function in patients with non-alcoholic fatty liver disease (NAFLD) and explore the relationship of these alterations with cognition and clinical indicators.

Materials and methods

In this cross-sectional study, forty-three patients with pre-cirrhotic NAFLD (male: 37, mean age: 38.2 ± 6.7 years) and twenty-three age-, sex-, and education-matched controls (male: 17, mean age: 41.0 ± 6.7 years) underwent diffusion tensor imaging (DTI) examination and cognitive measurements. The DTI analysis along the perivascular space (DTI-ALPS) index, calculated from the DTI data, assessed differences in GS function between the two groups. Linear regression analysis examined the relationship between the ALPS index and Z-transformed cognitive scores. Spearman/Pearson correlation analysis was conducted for assessing the relationship of the ALPS index with clinical indicators.

Results

After adjusted for age, sex, and BMI, NAFLD patients exhibited significantly lower ALPS index and higher diffusivity of projection fibers in the direction of the y-axis than controls (both P < 0.001). In the NAFLD group, the ALPS index was significantly correlated with the Mini-Mental State Examination score (beta [95% CI] = 2.123 [0.156, 4.091], P = 0.035) and the clock drawing score (beta [95% CI] = 4.233 [0.073, 8.393], P = 0.046) after adjusting for age, sex, body mass index, and education level. In addition, there was a significantly positive correlation between the ALPS index and abdominal visceral adipose tissue area (r = 0.353, P = 0.020) after adjusting for age, sex, and BMI in NAFLD patients.

Conclusion

NAFLD patients without cirrhosis showed significantly a reduced ALPS index compare to matched controls, indicating potential alterations in GS function. These alterations correlated with cognitive performance and clinical indicators, implying a possible link between GS function and cognitive performance in NAFLD that requires longitudinal validation.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-07432-9.

Introduction

It has been established that non-alcoholic fatty liver disease (NAFLD) is associated with cognitive dysfunction in adults [1–3]. NAFLD may even elevate the risk of dementia due to concomitant liver and metabolic dysfunctions [1, 4]. Currently, the mechanisms underlying cognitive dysfunction in NAFLD patients remain unclear, and hypotheses include neuroinflammation, insulin resistance, hyperammonemia, endothelial dysfunction, and gut microbiota disorders [5]. Emerging neuroimaging evidence indicates that structural and functional brain alterations—such as reduced brain volume, diminished cerebral perfusion, and lower cerebral oxygen concentration—may also contribute to cognitive decline in these patients [2, 6, 7]. Interestingly, the theory of intracerebral waste removal may provide further insight into the mechanisms of cognitive dysfunction in patients with NAFLD [8].

The glymphatic system (GS) has recently been identified as a whole-brain paravascular network that facilitates macromolecular clearance and may contribute to neuroinflammatory and neurodegenerative disease pathogenesis [9, 10]. This highly organized fluid transport system enables directional cerebrospinal fluid movement, whereby metabolic waste products are eliminated from the brain through perivascular spaces (PVS) and subsequently drained via meningeal and peripheral lymphatic vessels [11]. Efficient GS function is crucial for metabolic waste clearance, while its impairment has been implicated in the progression of neurocognitive disorders including Parkinson’s disease (PD) and Alzheimer’s disease (AD) [8, 12, 13]. In addition, emerging evidence suggests GS dysfunction also contributes to hepatic encephalopathy (HE) pathogenesis [14–16]. For example, studies in cirrhotic rats with HE demonstrated an impaired GS function in multiple brain regions (olfactory bulb, prefrontal cortex, and hippocampus), correlating with cognitive deficits [15]. Notably, Hsu et al. found that the improvement of meningeal lymphatic function facilitates glymphatic circulation, enhancing waste removal and thereby alleviating HE [14]. While HE represents an advanced cirrhotic complication with severe neurological manifestations [17], similar neuroinflammatory processes and hyperammonemia occur in pre-cirrhotic NAFLD patients [18, 19]. GS dysfunction is known to cause insufficient removal of inflammatory factors and ammonia from the brain [16]. To date, it remains unclear whether GS dysfunction occurs in patients with pre-cirrhotic NAFLD.

Intrathecal gadolinium-enhanced magnetic resonance imaging (MRI) has traditionally been regarded as the gold standard for evaluating glymphatic clearance function. However, this invasive approach requires contrast administration and repeated scanning, posing significant practical limitations [20]. Recently, a new noninvasive method called diffusion tensor image analysis along the perivascular space (DTI-ALPS) index has emerged as a promising alternative for evaluating GS function. DTI-ALPS utilizes diffusion tensor imaging to compute the ALPS index [21]. This method calculates the ALPS index by quantifying water molecule diffusivity along perivascular spaces, eliminating the need for contrast injection [22]. The current study has shown a high agreement between the noninvasive ALPS index and conventional glymphatic measurements performed with intrathecal gadolinium injection [23], validating its potential as a reliable neuroimaging biomarker of glymphatic function. The ALPS index has been extended to various diseases such as PD, AD, and cerebral small-vessel disease [12, 23, 24],, with ALPS index alterations showing significant associations with cognitive performance in these conditions [12, 21, 24]. These findings support the utility of DTI-ALPS for investigating cognitive impairment mechanisms in neurological disorders.

The purpose of this study was to evaluate GS function in patients with pre-cirrhotic NAFLD using the ALPS index compared to age-, sex-, and education-matched controls, and establish relationship between the ALPS index and cognitive scores. We hypothesized that NAFLD patients have significant GS alterations prior to hepatic cirrhosis and that the GS function correlates with cognitive performance.

Materials and methods

This cross-sectional study was conducted at the First Affiliated Hospital (Ethics Approval No. 2016 − 246) and the Second Affiliated Hospital of Wenzhou Medical University (Ethics Approval No. 2017-KY-83). The study protocol was approved by the ethics committees of these two hospitals. Written informed consent was obtained from all participants prior to the start of the study process.

Participants

Between July 2017 And May 2019, a total of 66 participants, including 43 biopsy-proven pre-cirrhotic NAFLD (37 male, mean age 38.2 years) And 23 age-, sex-, and education-matched controls (17 male, mean age 41.0 years) were prospectively recruited from both hospitals. The diagnosis of pre-cirrhotic NAFLD was based on the Liver histopathological examination. The pre-cirrhosis was determined by fibrosis stage less than 4 (F < 4) [25]. The controls were defined as hepatic MRI-Proton Density Fat Fraction (MRI-PDFF) < 6.4% (fatty liver was diagnosed when MRI-PDFF ≥ 6.4% [26]) and no abnormal lesion on conventional abdominal MR image. Participants were excluded if they met the following criteria: (1) left-handed or mixed-handed individual; (2) prior history of psychiatric disorders; (3) history of excessive alcohol abuse (consuming ≥ 14 alcoholic drinks per week for men or ≥ 7 alcoholic drinks per week for women) [2]; (4) history of type 2 diabetes [27]; and (5) poor imaging quality or abnormal lesion on brain MRI.

Neurocognitive function was evaluated by a blinded neurologist using standardized assessments: the Mini-Mental State Examination (MMSE) for global cognition, Clock Drawing Test (CDT) for visuospatial abilities, Digit Span Test (DST-Forward and DST-Backward) to assess working memory and executive function, and Rey-Osterrieth Complex Figure Test (CFT-Copy and CFT-Delay) for working memory and executive processing [7, 28]. A detailed overview of the cognitive assessment is shown in Table S1. Blood biochemistry tests included total bilirubin, fasting glucose, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), γ-Glutamyltransferase (γ-GT), total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDL-cholesterol), and low-density lipoprotein cholesterol (LDL-cholesterol). Neuropsychological assessments and laboratory examinations were performed on the same day as the MRI scan.

Histopathological evaluation

All patients with NAFLD underwent ultrasound-guided percutaneous biopsy of the right lobe of the liver. The detailed assessment processes of liver pathology specimens are shown in Supplementary E1.

ALPS index acquisition

The DTI data of all participants were acquired by using a 3.0T MRI scanner (DISCOVERY MR750, General Electric Healthcare, USA) with an 8-channel phased array head coil. The DTI examination was performed by a spin-echo Single-shot echo planar imaging sequence with 64 different diffusion directions (TR/TE of 7000/61ms, FA of 90°, acquisition matrix of 128 × 128, FOV of 256 × 256 mm², b values of 1000 And 0 s/mm², number of excitation 1, without gap, And 69 axial slices at 2 × 2 × 2 mm resolution). All participants were required to keep awake to minimize the effects of sleep states on GS [29].

ALPS index calculation followed these standardized steps: First, DTI DICOM files were converted to NIFTI format using MRIcroGL GUI. We then implemented an automated processing pipeline combining FSL and MRtrix3 tools to compute the ALPS index from DTI data [22]. FSL was employed to generate the FA and diffusivity maps along the x-, y-, and z-axes. Each subject’s FA map was co-registered to the JHU-ICBM FA template, and the resulting transformation matrix was applied to all diffusivity maps. Based on the JHU-ICBM-DTI-81-whitematter Labeled Atlas, the superior corona radiata (SCR) and superior longitudinal fasciculus (SLF) fibers were identified. The regions of interest (ROIs) (5 mm in diameter) were placed in the bilateral SCR and SLF areas and this definition was applied to all subjects’ diffusivity maps. The mean diffusivities of the projection fiber in direction of x- and y-axes (Dxxproj and Dyyproj) and the association fiber in direction of x- and z-axes (Dxxassoci and Dzzassoci) were extracted to calculate the ALPS index, which partially represented water diffusivity along the PVS (Fig. 1A and B). The mean of bilateral ALPS indexes was considered as the true ALPS index. In addition, the ALPS index was calculated to estimate GS activity for all participants using the following formula:

Fig. 1.

Analytical process for calculating the ALPS index. The yellow points in the color FA map represent the ROIs corresponding to projection and association fibers areas, respectively (A). These ROIs are then applied to the diffusivity maps. Subsequently, the mean diffusivities of the projection fibers along the x- and y-axes and the association fibers along the x- and z-axes are extracted to calculate the ALPS index (B). ALPS, diffusion tensor image analysis along the perivascular space; FA, fractional anisotropy; ROIs, regions of interest

MRI-PDFF measurements

A 3D iterative decomposition of water and fat with echo asymmetry and least squares estimation quantification (IDEAL-IQ) technology was used for quantitative MRI-PDFF evaluation [30]. Using this sequence, liver fat fraction (MRI-PDFF) was evaluated. In addition, the area of abdominal visceral adipose tissue (VAT) at the level of the third lumbar vertebra was also measured. Given that VAT is closely associated with various metabolic and cardiovascular diseases, especially NAFLD, and may also have an impact on the brain [31, 32]. Detailed sequence parameters and the methodology for measuring derived parameters were provided in Supplementary E2.

Statistical analysis

Continuous variables were compared between groups using either the independent samples t-test (for normally distributed data) or the Mann-Whitney U test (for non-normally distributed data), with normality assessed by the Kolmogorov-Smirnov test. Categorical variables were analyzed using the chi-square test. Mean diffusivity along the x-, y-, and z-axes directions and the ALPS index between groups were compared using analysis of covariance (ANCOVA) with age, sex, and BMI adjusted. The ANCOVA was also used to compare the difference of cognitive score between groups, with age and educational level adjusted. The reason for controlling for these covariates was to rule out their potential effect on cognition and GS [33]. Continuous variables were expressed as means ± standard deviations or medians with interquartile ranges, while categorical variables were expressed as a percentage of the data.

Raw cognitive scores (as shown in Table S2) were converted to standardized z-score. A higher z-score represents better cognitive performance, and a lower z-score represents a worse cognitive performance. Linear regression models were used to assesses the relationship between the ALPS index and Z-transformed cognitive scores, adjusted for potential confounding factors, age, sex, BMI (to account for obesity-related neurological effects) [34], and educational level. The relationships of the ALPS index with vascular risk factors (VAT, blood pressure, fasting glucose, and lipid metabolism-related indicators), live MRI-PDFF, and nonalcoholic steatohepatitis (NASH) status were investigated using Spearman/Pearson correlation (as appropriate) analysis with age, sex, and BMI controlled. A two-sided p value less than 0.05 was statistically Significant. SPSS 27.0 software was used for all statistical analyses.

Results

Clinical characteristics and cognitive performance differences

The clinical characteristics and Z-transformed cognitive scores of the two groups are shown in Table 1. The age, sex, and education level of patients with NAFLD were not significantly different from those in controls. However, significantly higher BMI, ALT, AST, ALP, γ-GT, TC, TG, LDL-cholesterol, and lower HDL-cholesterol were found in patients with NAFLD compared to the controls (all p < 0.05). Cognitive scores did not differ significantly between the two groups.

Table 1.

Demographic and clinical characteristics in patients with NAFLD and the controls

	Pre-cirrhosis NAFLD (n = 43)	Control (n = 23)	p-value
Clinical data
Age, years	38.2 ± 6.7	41.0 ± 6.7	0.116
Education, years	9 (9–12)	12 (9–16)	0.054
Male, n (%)	37 (86.0)	17 (74.0)	0.319
BMI, kg/m2	26.62 (24.77–29.37)	24.69 (20.90-26.57)	0.002
Hypertension, n (%)	18 (41.9)	5 (21.7)	0.102
Systolic pressure, mmHg	129.77 ± 17.45	123.39 ± 11.22	0.118
Diastolic pressure, mmHg	84 (74–92)	80 (71–82)	0.128
Laboratory indictors
Total bilirubin, µmol/L	15 (12–22)	14 (10–19)	0.278
Fasting glucose, mmol/L	5.0 (4.8–5.5)	4.9 (4.6–5.4)	0.230
Alanine aminotransferase, U/L	47 (32–87)	29 (20–62)	0.025
Aspartate aminotransferase, U/L	32 (26–50)	24 (20–35)	0.005
Alkaline phosphatase, U/L	80.84 ± 17.36	70.43 ± 21.49	0.037
γ-Glutamyltransferase, U/L	55 (36–85)	31 (15–56)	0.004
Total cholesterol, mmol/L	5.04 ± 1.09	4.44 ± 0.92	0.028
Tryglicerides, mmol/L	1.92 (1.32–2.91)	1.21 (0.76–1.88)	0.004
HDL-cholesterol, mmol/L	1.04 (0.89–1.11)	1.13 (1.01–1.22)	0.003
LDL-cholesterol, mmol/L	3.14 ± 0.86	2.66 ± 0.76	0.026
Hepatic Pathology and Abdominal adipose tissue
Visceral adipose tissue, mm2	17052.94 (14318.51–22372.00)	11740.57 (7065.71-21221.26)	0.002
Hepatic MRI-PDFF, %	9.57 (6.54–13.01)	3.71 (2.29–4.86)	< 0.001
NASH disease status, n (%)
NAFL	8 (18.6)
Borderline NASH	18 (41.9)
Define NASH	17 (39.5)
Cognitive measurement (Z -transformed)
Mini-Mental State Examination	−0.18 ± 0.51	0.31 ± 1.53	0.267
Clock Drawing Test	−0.05 ± 1.03	0.07 ± 0.98	0.866
Complex figure test-Copy	0.01 ± 1.00	−0.01 ± 1.03	0.692
Complex figure test-Delay	0.15 ± 1.07	−0.27 ± 0.84	0.053
Digital Span Test-Forward	0.11 ± 0.76	−0.19 ± 1.35	0.235
Digital Span Test-Backward	−0.04 ± 1.06	0.11 ± 0.90	0.606

NAFLD Non-alcoholic fatty liver disease, BMI Body mass index, HDL High-density lipoprotein, LDL Low-density lipoprotein, MRI-PDFF Magnetic resonance imaging-proton density fat fraction, NAFL Non-alcoholic fatty liver, NASH Nonalcoholic steatohepatitis

The NAFLD group had significantly higher liver MRI-PDFF and VAT (all p < 0.001) than the control group. Based on the NAS, patients were categorized into the following categories: non-NASH (n = 8), borderline NASH (n = 18), and definite NASH (n = 17) (Table 1).

ALPS index differences

Compared with the controls, NAFLD patients showed significantly lower ALPS index (1.38 ± 0.09 vs. 1.46 ± 0.12, p < 0.001) (Fig. 2A). The NAFLD group exhibited a significantly higher diffusivity of projection fibers in the direction of the y-axis than the controls (0.00055 ± 0.00005 vs. 0.00050 ± 0.00005, p < 0.001) (Fig. 2B). There was no significant difference in the diffusivities of projection fibers in the direction of the x-axis and the association fibers in the direction of x- and z-axes between the two groups. A summary of the detailed results is provided in Table 2.

Fig. 2.

Violin plots display the distribution of the ALPS index (A) and Dyyproj (B) between NAFLD and control groups. The Lines correspond to the 25th percentile, median, And 75th percentile. *** represents a p value < 0.001. ALPS, diffusion tensor image analysis along the perivascular space; Dyyproj, diffusivities of the projection fiber in direction of y-axis

Table 2.

Differences of the diffusivities and ALPS index in patients with NAFLD and the controls

	Pre-cirrhosis NAFLD (n = 43)	Control (n = 23)	p-value
Mean Dxxproj (×10−3)	0.61 ± 0.03	0.62 ± 0.03	0.309
Mean Dxxassoc (×10−3)	0.70 ± 0.04	0.70 ± 0.04	0.489
Mean Dyyproj (×10−3)	0.55 ± 0.05	0.50 ± 0.05	< 0.001
Mean Dzzassoc (×10−3)	0.40 ± 0.03	0.40 ± 0.03	0.886
Mean ALPS index	1.38 ± 0.09	1.46 ± 0.12	< 0.001

The “Mean” represents the mean of the diffusivity measurements from the two hemispheres

NAFLD Non-alcoholic fatty liver disease, DTI-ALPS Diffusion tensor image analysis along the perivascular space, Dxxproj Diffusivity along the x-axis in project fiber, Dxxassoc Diffusivity along the x-axis in association fiber, Dyyproj Diffusivity along the y-axis in project fiber, Dzzassoc Diffusivity along the z-axis in association fiber

Correlation analysis

In the NAFLD group, the ALPS index was significantly correlated with the MMSE (beta (95% CI) = 2.123 (0.156, 4.091), p = 0.035) and CDT scores (beta (95% CI) = 4.233 (0.073, 8.393), p = 0.046) (Table S3, Fig. 3A). No other significant correlations between the ALPS index and cognitive scores were observed in the two groups. There was a significantly positive correlation between the ALPS index and VAT (r = 0.353, p = 0.020) after adjusting for age, sex, and BMI in NAFLD patients (Table S4, Fig. 3B). However, there was no significant correlations between the ALPS index and other risk factors (blood pressure, fasting glucose, and lipid metabolism-related indicators) in either group (Table S4). No significant correlation of ALPS index with liver MRI-PDFF was found in the two groups (Table S4). The ALPS index was not correlated with NASH status in the NAFLD group (Table S4).

Fig. 3.

A The association of the ALPS index with cognitive scores in NAFLD group. Results are acquired using Linear regression Analysis with the age, sex, BMI, And educational level. A 95% confidence interval (CI) excluding zero is considered as significant different. * represents a p value < 0.05. B Relationship between the ALPS index and visceral adipose tissue in NAFLD and control groups. The X and Y axes represent the residual of the ALPS index and visceral adipose tissue after regressing out the effect of age, sex, and BMI, respectively. ALPS, diffusion tensor image analysis along the perivascular space

Discussion

This pilot study evaluated the brain glymphatic circulation function in NAFLD patients and found a significant decrease in the ALPS index in patients with pre-cirrhosis NAFLD compared to the controls. In addition, the ALPS index was significantly correlated with cognitive scores in the NAFLD group.

It is well-established that the ALPS index can be used to assess GS function. A previous study reported dysfunctional GS in cirrhotic rats with HE [15]. Similarly, our study found a significantly lower ALPS index in patients with pre-cirrhotic NAFLD compared to the controls, with controlling for confounding factors. Our results suggested that GS was impaired even in the pre-cirrhotic stage of NAFLD, which had not been reported in the previous literature. While a recent study reported altered ALPS index in pre-cirrhotic metabolic dysfunction-associated fatty liver disease patients [35], the pathophysiological distinctions between NAFLD and MAFLD suggest these findings may not be fully generalizable. Nevertheless, both studies indicate that glymphatic dysfunction may occur early in the course of fatty liver disease, prior to cirrhosis development. The cerebral glymphatic circulation is reportedly affected by many anatomical and physiological factors, such as arterial pulsation, respiratory effort, sleep, CSF pressure gradient, and aquaporin-4 (AQP4) water channels [8]. Current evidence suggests that dysfunctional GS was mediated by decreased AQP4 expression in cirrhotic rat brains [15]. Decreased expression levels of AQP4 attenuate efficient glymphatic influx and efflux, leading to the deposition of metabolic waste products. Although the reduction in AQP4 polarization did not reach significance in that model [15], its established role in neurodegenerative diseases—which share neuroinflammatory and metabolic features with NAFLD—suggests that mislocalization may also contribute to glymphatic dysfunction in our cohort [1, 36, 37]. Further studies are needed to determine whether loss of AQP4 expression, loss of polarization, or both underlie glymphatic impairment in NAFLD. It has been shown that NAFLD leads to vascular endothelial dysfunction and arterial stiffness, which attenuate arterial wall pulsatility and compliance, resulting in increased blood pressure [38, 39]. In our study, more than 40.0% of NAFLD patients had hypertension (systolic pressure above 140mmHg or diastolic pressure above 90mmHg), with a prevalence twice as high as in the control group. Mestre et al. reported that increased blood pressure in live mice may reduce net CSF flow in PVS and lead to a decline in glymphatic circulation, demonstrating the role of blood pressure in the glymphatic dysfunction [40]. However, our study found no significant correlation between blood pressure and the ALPS index, suggesting that glymphatic function may not be directly associated with blood pressure regulation in early-stage NAFLD. This is inconsistent with the findings of Mestre et al. We hypothesize that this may be due to the fact that not all patients with NAFLD suffer from hypertension, or it may be due to the inherent heterogeneity between humans and rats. Hence, the role of blood pressure in the development of glymphatic dysfunction in patients with NAFLD still remains controversial, which requires further research. NASH can affect the GS function and results in the loss of neurons, infiltration of lymphocytes, and increased activation of microglia and astrocytes [1, 41]. NASH-induced reactive changes in microglia and astrocyte morphology influence glymphatic circulation and reduce the clearance of inflammatory cytokines [42, 43]. More than 80.0% of NAFLD patients in the presented study with NASH or borderline NASH. Although impaired GS (decreased ALPS index) was found in NAFLD patients, the ALPS index was not correlated with NASH status. This may be attributed to the inclusion of patients with borderline NASH, as well as the uncertainty surrounding the determination of NASH status, which could have obscured the relationship between NASH and the ALPS index. Despite the presence of the above-mentioned factors influencing GS function in NAFLD, there is a scarcity of studies exploring the underlying mechanisms of GS disorders in this condition. Further studies are needed to demonstrate the exact mechanisms by which these factors lead to GS disorders.

Our results demonstrated that the reduced ALPS index showed positive associations with MMSE and CDT scores in NAFLD patients after adjusting for covariates (age, sex, BMI, and education level), consistent with previous findings [44]. Notably, no significant difference in MMSE and CDT scores was found between the two groups since most patients in this study had mild-to-moderate fibrosis (F ≤ 2, n = 42). A previous longitudinal study consistently showed no significant change in cognitive performance in NAFLD patients compared to the controls, and this cognitive change may be related to the duration of NAFLD [45]. Considering that the association between the ALPS index and cognitive scores was only observed in the NAFLD group and not in the control group, it can be inferred that this relationship is unique to NAFLD, and alteration of glymphatic function may affect cognitive performance in patients with NAFLD. Therefore, the ALPS index may serve as a potential early marker of cognitive alterations in NAFLD patients, which is required further verified.

Higher VAT was found in the NAFLD group compared to the control group. Excessive VAT deposition remained significantly associated with the ALPS index in NAFLD after adjusting for age, sex, and BMI, suggesting a potential influence of VAT on glymphatic function. Recent studies have demonstrated that NAFLD is related to the development and progression of AD [46]. One hundred and eighty-nine genes co-expressed in NAFLD and AD have been identified; among these, low-density Lipoprotein receptor-related protein 1 has been reported to promote cerebral amyloid-β (Aβ) clearance and co-expression in the liver and BBB [47]. In addition, VAT metabolism was strongly associated with cerebral Aβ burden and contributed to AD progression [48]. A significant association was found between the ALPS index and CSF Aβ levels in AD, suggesting the importance of perivascular clearance functional integrity for Aβ clearance from the brain [49]. Based on these findings, we hypothesized that impaired GS in NAFLD may reduce the clearance of CSF Aβ, which may accelerate cognitive impairment. However, further research is warranted to verify whether there is high Aβ deposition in NAFLD and whether the ALPS index influences its metabolism.

Several limitations warrant consideration. First, while the ALPS index shows high consistency with gadolinium-based glymphatic MRI, its pathophysiological link to human glymphatic function remains unestablished [23]. Animal studies are needed to verify these findings in pre-cirrhotic NAFLD models and assess cognitive implications. Second, our evaluation focused on periventricular areas, which may not reflect whole-brain glymphatic circulation. Third, although we standardized wakefulness during MRI to avoid sleep-state confounders [8], unmeasured sleep quality differences between groups could influence glymphatic function. Fourth, while BMI was included as a covariate in statistical analyses, residual confounding from baseline group differences remains possible. Additionally, sex-stratified analyses were precluded by sample size limitations, representing an important consideration for future studies. Finally, the single-center cross-sectional design with limited sample size restricts statistical power and causal inference.

Conclusions

In this study, a decreased ALPS index was revealed in pre-cirrhotic NAFLD patients, which may reflect potential alterations in glymphatic function before cirrhosis onset. The observed association between ALPS index changes and cognitive performance suggests a possible link between glymphatic function and cognitive performance in NAFLD. While the cross-sectional design of this study limits causal interpretation, these preliminary findings warrant further investigation through longitudinal studies with larger NAFLD cohorts.

Authors’ contributions

Author contribution: Kun Shu: Conceptualization, Formal analysis, Visualization, Writing-original draft; Shaoqing Chen: Conceptualization, Writing-original draft; Shuang Meng: Formal analysis, Writing-review & editing; Yang Yang: Investigation, Validation & Data curation; Jiawen Song: Investigation, Validation; Xiaoyan Huang: Data curation, Validation, Writing-review & editing; Xinjian Ye: Investigation, Writing-review & editing; Shihan Cui: Data curation, Writing-review & editing; Yongjin Zhou: Data curation, Writing-review & editing; Lu Han: Methodology, Software, Writing-review & editing; Peng Wu: Methodology, Software, Writing-review & editing; Zhihan Yan: Conceptualization, Project administration, Resources, Supervision, Writing-review & editing; Kun Liu: Conceptualization, Project administration, Resources, Supervision, Writing-review & editing, Data curation.

Funding

This work was supported by the grants from Zhejiang Provincial Natural Science Foundation (LY18H070003 and LY19H180003), National Natural Science Foundation of China (82071902), and Wenzhou Science and Technology Bureau (No. Y20220066).

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to confidentiality but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval for the study was obtained from the ethics committee of the First Affiliated Hospital and the Second Affiliated Hospital of Wenzhou Medical University. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Informed consent was obtained from all patients for being included in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.