Cystatin C is highly expressed in the brain and cerebrospinal fluid (CSF) (Löfberg and Grubb, 1979), and it is indicated to play many roles in the risk and pathobiology for Alzheimer's disease (AD) (Taupin et al., 2000; Palmer et al., 2001; Gauthier et al., 2011). Genetically, the CST3 B haplotype of cystatin C was considered to be a risk factor for AD, frontotemporal dementia (FTD), and Lewy body dementia (LBD) (Finckh et al., 2000; Bertram et al., 2007; Maetzler et al., 2010; Hua et al., 2012). Clinically, in patients with FTD and LBD, a reduction of CSF cystatin C was found, and it was associated with an anticipation of dementia onset (Rüetschi et al., 2005; Sundelöf et al., 2008; Maetzler et al., 2010).

Specifically, in patients with AD, cystatin C levels in CSF were found to be reduced, compared to individuals without dementia (Simonsen et al., 2007; Hansson et al., 2009; Zhong et al., 2013), and CSF cystatin C levels were positively correlated with tau and Aβ levels (Sundelöf et al., 2010; Zhong et al., 2013). However, another study found increased cystatin C levels in CSF of AD patients compared to those in controls (Carrette et al., 2003). In plasma, a study found that AD patients had higher plasma cystatin C levels than healthy control subjects (Wang R. et al., 2017). The association of serum cystatin C with risk of mild cognitive impairment (MCI) or dementia was inconsistent. The Health ABC Study showed that high levels of serum cystatin C increased the risk of cognitive impairment and individuals with higher levels of cystatin C had poorer performance on cognitive examinations (Yaffe et al., 2008). In contrast, the Uppsala Longitudinal Study found that high levels of serum cystatin C were related to a decreased risk of AD in men aged between 70 and 77 years-old (Sundelöf et al., 2008), and a study also found that MCI patients with higher serum cystatin C levels remained stable, without developing to dementia (Romero-Sevilla et al., 2018). Moreover, the Osteoporotic Fractures study reported a U-shape association between serum cystatin C and cognitive impairment in elderly women, but the association no longer existed after adjusting for covariates (Slinin et al., 2015). However, there is no study focusing on the association of serum cystatin C levels with cognitive performance in AD patients.

In this study we are aiming to examine the levels of serum cystatin C in AD patients and compare to those in matched healthy subjects, and trying to determine whether serum cystatin C levels are associated with the severity of the dementia.

This cross-sectional study included 463 AD patients [231 males (49.9%) and 232 (50.1%) females], and 1,389 healthy subjects [693 males (49.9%) and 696 females (50.1%)]. The MRI scan supported the diagnosis of the included AD patients, who had medial temporal 146 lobe atrophy, and the Fazekas scale of white matter lesions was <2. The mean age at examination of the AD patients and HCs were 69.00 ± 11.31 and 69.08 ± 11.28 years, respectively. Serum cystatin C levels in AD patients were significantly higher than those in normal subjects (1.034 ± 0.254 vs. 1.010 ± 0.248, p = 0.0362). The cognitive performances evaluated by MMSE or MoCA were poorer in AD patients compared to those in HCs (Table 1).

In order to determine the association between serum cystatin C levels and cognitive impairment, we stratified the cohort based on the CDR grade (mild CDR = 0.5 or 1, moderate CDR = 2, severe CDR = 3). The one-way ANOVA analyses showed that the serum cystatin C levels were not significantly different among AD patients with different severities of dementia (p = 0.6588). In addition, the correlation analyses were utilized to investigate the correlations between serum cystatin C levels and cognitive performance assessed by MMSE and MoCA (Table 2), and a significant correlation between serum cystatin C levels and MoCA scores was identified (r_s = −0.1326, p = 0.046). However, the correlation analyses showed that serum cystatin C levels were not correlated to the cognitive performance in HCs (MMSE: r_s = −0.09827, p = 0.1728; MoCA: r_s = −0.1455, p = 0.0798). Serum cystatin C levels was also found to be significantly associated with age at examination in both AD patients and HCs (AD patients: r_s = 0.4869, p < 0.0001; HCs: r_s = 0.5033, p < 0.0001). The ordinal logistic regression stratified by CDR grade was conducted, and the results of the model for predicting CDR grade using cystatin C level showed age at onset, sex, disease duration and presence of APOE ε4 allele were not significant risk factors in this model. However, serum cystatin C levels were slightly associated with severity of cognitive impairment (OR = 1.438, 95% CI 0.017–2.858, p = 0.047). With regard to the cognitive predictors, AD patients with a 1-point increase in cystatin C levels were 48% more likely to have a higher CDR grade, indicating the AD patients with higher cystatin C levels may have increased odds of severe cognitive dysfunction.

Cystatin C, a cysteine protease inhibitor, is produced by most nucleated cells and present in all body fluids. In the present study, we found that serum cystatin C levels were increased in AD patients than healthy controls, which was supported by some studies (Straface et al., 2005; Wang R. et al., 2017). In contrast, a previous study reported that the plasma cystatin C levels were lower in AD patients than in controls (Chuo et al., 2007), and another studies did not find significant differences in plasma cystatin C levels between AD patients and healthy controls (Kálmán et al., 2000; Ghidoni et al., 2010; Zhong et al., 2013). Plasma cystatin C was also indicated to modulate the clinical expression of cognitive decline; a significant anticipation of the conversion to dementia was observed in MCI subjects, when the detected plasma cystatin C levels were below 1,067 ng/ml (Ghidoni et al., 2010). Such discrepancy may be caused by the differences in sample sizes, characters of participants and genetic backgrounds. To our knowledge, the current study included the largest sample to determine the difference in serum cystatin C levels between AD patients and healthy subjects.

Cystatin C is an inhibitor of cathepsins. Cathepsin D is suggested to be involved in the pathogenesis of AD, and the cathepsin D level is increased in AD patients (Nixon, 2000). The increase of this cathepsin-inhibitory enzyme in AD patients may represent a compensatory activity aimed at counteracting the increased presence of cathepsin D. Cystatin C levels were also increased in response to injury and oxidative stress (Finckh et al., 2000; Nishio et al., 2000). In the central nervous system (CNS), cystatin C can protect neuronal cells from degeneration induced by fibrillar and oligomeric Aβ (Tizon et al., 2010). Co-incubate cystatin C with monomeric Aβ42 can attenuate the formation of Aβ oligomers and protofibrils (Sastre et al., 2004; Selenica et al., 2007), and increase the cystatin C expression can reduce parenchymal Aβ load in CNS (Kaeser et al., 2007; Mi et al., 2007). In addition, the peripheral production of Aβ can be derived from peripheral organs and tissues, including various blood and endothelial cells (Wang J. et al., 2017). The accumulation of Aβ deposits in the peripheral system may stimulate the expression of cystatin C. However, the exact mechanism of increased serum cystatin C in AD patients is not clear. We have to notice that the serum cystatin C is cleared from the circulation by glomerular filtration, and it can be affected by renal function, but the increase of serum cystatin C level in AD patients cannot be simply attributable to the changes of renal function since we did not find any difference in eGFR between AD patients and controls. Furthermore, ANCOVA adjusting for eGFR showed consistent result of significantly higher serum cystatin C levels in AD patients. The findings of the association of serum cystatin C with cognitive impairment were inconsistent: one study did not find any associations in AD patients (Zhong et al., 2013), but another study found a significant correlation between serum cystatin C levels and MMSE scores in female AD patients (Wang R. et al., 2017). In the current study, we found that serum cystatin C levels were negatively associated with cognitive function assessed by MoCA, higher serum cystatin C levels were weakly related to poorer cognitive performance. In addition, we found a significant positive correlation between serum cystatin C and age in both AD patients and healthy subjects. Previous studies also found this association in AD patients (Chuo et al., 2007; Zhong et al., 2013). Therefore, the association between serum cystatin C and age is not disease-specific. Since age is the biggest risk factor for AD (McCartney et al., 2018), we compared the age-at-examination among AD patients with different CDR score, but the differences in age-at-examination were not significant in patients with different level of cognitive dysfunction. We also compared the serum cystatin C levels among AD patients with different severity of dementia, and we did not find any differences in serum cystatin C levels among patients with mild, moderate, or severe dementia evaluated by CDR. However, using ordinal logistic regression models, we found that serum cystatin C levels were slightly associated with cognitive dysfunction; patients with higher cystatin C levels had increased odds of severe cognitive impairment. Similarly, a previous study found that among community-resident elders, those with increased serum cystatin C levels had poorer performance in cognitive tests evaluated by the Modified Mini-Mental State Examination (3MS) and the Digit Symbol Substitution Test (DSST) (Yaffe et al., 2008). Another study also evaluated the cystatin C levels in 193 participants older than 90, and found that higher tertiles of cystatin C was related with poorer global cognition, executive function and visual-spatial ability (Lau et al., 2020). Similarly, the Cardiovascular Health Study Cognition Study reported that high serum levels of cystatin C were related to poorer cognitive performance 6 years later (Riverol et al., 2015). However, the results of previous studies that correlated cystatin C with AD required careful interpretation. First, the association between cystatin C and cognition may underscore the connection between kidney function and cognitive function. Studies have shown that subjects with renal dysfunction had an elevated risk of developing dementia and poorer performance on cognitive function (Seliger et al., 2004; Martens et al., 2017). Cystatin C was found to be associated with cognitive performance in 90+ year-olds; among them nearly 90% had cystatin C levels more than 1.0, indicating that the kidney function is impaired. Further study also gave 308 individuals aged 90 or older a PET scan and obtained the brain indices of Aβ deposition using a statistically defined region of interest (statROI), and found that PET statROI was not correlated with cystatin C and eGFR, indicating an independent association between cognition and chronic kidney disease (CKD), and the CKD-associated cognitive dysfunction largely reflects vascular rather than Aβ pathology (Lau et al., 2021). Second, the correlation between serum cystatin C and cognitive function may not only depend on kidney function. In the present study, no significant differences in eGFR were found among patients with different degrees of dementia stratified by CDR grade, and the eGFR values were not associated with cognition evaluated by MMSE and MoCA. In addition, a previous study also found that elders with high serum cystatin C levels had higher risks of cognitive impairment whether or not they had CKD (Yaffe et al., 2008).

There are some limitations in the present study which should not be ignored: (1) the genetic CST3 genotypes were not considered; (2) the results of the present cannot be extrapolated to other populations since our patients were recruited from a single specialized unit. On the other hand, some of the strengths of our study are the strict inclusion criteria, the examination of APOE genotype, and the fact that the specialized neurologists conducted all the data collections and manifestation assessments.

In summary, the findings of the current study support the idea that cystatin C plays an essential role in the pathogenesis of AD. Our results suggest AD patients had higher levels of serum cystatin C when compared to age/sex/BMI-matched normal controls, and higher serum cystatin C levels were weakly associated with worse cognitive performance in AD patients. Our findings strengthen the evidence that serum cystatin C may be involved in modulating the clinical expression of cognitive decline. Further longitudinal studies in larger cohorts and distinct ethnic groups will be needed to validate our findings, and to determine the mechanisms underlying this association.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

HS conceived this study. YH, TB, RO, QW, YC, JL, and JY did the patient evaluation and data collection. FJ did the statistical analysis. XC and HS wrote the manuscript while all authors revised and discussed to the final edition. All authors contributed to the article and approved the submitted version.

This study was supported by grant from National key Research and development program of China (Grant No. 2017YFC09007703), grant from science and technology planning project in Sichuan Province (Grant No. 2020YJ0281), grant from 1·3·5 project for disciplines of excellence West China Hospital Sichuan University (Grant No. ZYJC18038), and the grant from cadres health care project in Sichuan Province (Grant No. 2019-112).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.