This systematic review and meta-analysis used the Priority Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. From the earliest available date to December 1, 2022, we conducted electronic literature searches using PubMed, Embase, Web of Science, and the Cochrane Library. Literature searches were conducted using the following search terms/keywords, but not limited to: (“Lewy body disease” OR “Lewy Body Dementia” OR “Diffuse Lewy Body Disease” OR“Cortical Lewy Body Disease” OR “Lewy Body Disease, Cortical” OR “Lewy Body Disease, Diffuse” OR “Lewy Body Type Senile Dementia” OR “Dementia, Lewy Body”) and (“albumin CSF/serum ratio” OR“cerebrospinal fluid/serum albumin ratio” OR “value of albumin quotients” OR “Qalb”). The searches were not restricted by age, race, or region. In addition, references in retrieved papers and reviews were manually searched for eligible studies. Two authors independently conducted literature searches and screened all titles, abstracts, and full texts of eligible studies. Disagreements were discussed with a third reviewer and resolved by mutual agreement.

The search and screening strategy and eligibility criteria based on the PRISMA guidelines are shown as (Figure 1).

Our analyses included studies that met the following criteria: (1) the primary eligibility criterion was to compare Qalb levels in patients with LBD and healthy controls/any other dementia subtype; (2) studies that provided a complete definition of participants. Diagnoses of different types of dementia/healthy controls were determined using clinical assessments and neuropsychological tests: AD was defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; International Statistical Classification of Diseases and Related Health Problems, 10th Edition, National Institute of Neurological Communicative Disorders and Stroke Criteria, and AD and Associated Disorders Association Criteria (McKhann et al., 2011). PDD was defined by the Parkinson’s Disease Society United Kingdom Brain Bank definition (Hughes et al., 1992). DLB was defined according to the DLB consensus criteria (McKeith et al., 2017). FTLD was defined according to the International FTLD Consensus Criteria (Neary et al., 1998); and (3) Qalb levels were measured and sufficient data (pg/dl) were available. (4) Language is limited to Chinese or English.

Similarly, studies were excluded if any of the following criteria were met: (1) there were no available Qalb data; (2) the publication was a letter, review, meta-analysis, or animal study; and/or (3) the Qalb assay methodology differed from the guidelines.

The quality of the included studies was assessed according to the Newcastle-Ottawa Scale (NOS) criteria (Stang, 2010), including selection (0–4 points), comparability (0–2 points) and exposure (0–3 points). Scores ranged from 0 ~ 9, with no <7 indicating a high quality study. Disagreements were resolved by consensus and by seeking input from a third reviewer.

Data were analyzed using Review Manager 5.3 software. Since Qalb levels were continuous, standardized mean differences (SMD) were calculated and used for statistical analysis. The range between median and quartiles was converted to mean and standard deviation (SD) according to Hozo et al. (2005).

The funnel plot indicated the bias of the study; a symmetrical funnel plot was free of bias, and an asymmetrical funnel plot meant that bias was present. The Z-test was used to assess whether the results of the studies were significant or not, and a p value of <0.05 was considered significant, and vice versa. Heterogeneity between studies was assessed by Q value and I², and the cut-off level of heterogeneity was set at 0.5. If p ≥ 0.05 or I² ≤ 50%, the data were analyzed using the fixed-effects model; if p < 0.05 or I² > 50%, the random-effects model was used. Subgroup analysis was used to explore inter-study heterogeneity, and sensitivity analysis was performed by removing studies one by one to explore the source of heterogeneity (Zhang et al., 2022).

We identified a total of 86 studies. After excluding duplicates (31 cases), a total of 55 studies were obtained. Of these studies, 1 study was excluded because it was a review (n = 1). Of the remaining 53 studies, 29 were excluded after reading the abstracts, whereas 13 studies were excluded after a complete review (Pirttila et al., 1994; Sjögren et al., 2000, 2002; Brettschneider et al., 2005; Nielsen et al., 2007; Boström et al., 2009; Janelidze et al., 2015, 2017; Llorens et al., 2015; Skillbäck et al., 2017; Bekris et al., 2018; Musaeus et al., 2020; Gan et al., 2023). These included 225 DLBs, 302 PDs, 1723 ADs, 190 FTDs, and 622 healthy controls(HCs). The age range of LBD patients is between 68 to 74 years, with males constituting 25 to 73 percent of the cases (Table 1).

To compare Qalb levels in LBD patients with HCs, 12 studies were included (508 LBD patients, 622 HCs). The studies showed high heterogeneity (I² = 98%, p < 0.00001), so a random-effects model was used to calculate the pooled SMD. The forest plot showed statistically significant higher mean Qalb levels in LBD compared with HCs (SMD: 1.86, 95% CI: 1.06 to 2.67, Z = 4.53, p < 0.00001; Figure 2A).

This part of the meta-analysis included 10 studies. The number of patients with LBD and AD was 384 and 1723, respectively. Study heterogeneity was high (p = 0.010, I² = 58%), so a random-effects model was used to calculate the pooled SMD. A forest plot showed that Qalb levels were significantly higher in patients with LBD compared to patients with AD (SMD: 1.13, 95% CI: 0.42 to 1.83, Z = 3.15, p = 0.002; Figure 2B).

Seven studies were included in this part of the meta-analysis. There are 297 and 190 patients with LBD and FTLD, respectively. Study heterogeneity was high (p = 0.03, I² = 57%). The forest plot showed that Qalb levels were significantly higher in patients with LBD compared with those with FTLD (SMD: 1.13, 95% CI,0.14 to 2.13, Z = 2.24, p = 0.03; Figure 2C).

In the present study, we compared Qalb in LBD and healthy controls. We observed that Qalb levels were significantly higher in patients with LBD, with a significant difference compared to healthy controls (p < 0.05).DLB disease and Parkinson’s disease are the two main forms of LBD, sharing common clinical and neuropathological features (Wakabayashi et al., 2013). Previous studies have published reports that Qalb in DLB is increased compared with the healthy controls (Janelidze et al., 2015). Several important cerebrospinal fluid biomarker studies have reported elevated Qalb in patients with early PD compared to the healthy controls (Janelidze et al., 2015; Liguori et al., 2017; Page et al., 2021). These studies are consistent with our analysis, which suggests that impairment of the BBB may be present in patients with LBD.

Neuroinflammation is an important factor in the onset and progression of most neurodegenerative diseases. Alterations in BBB permeability caused by the release of proinflammatory cytokines directly or indirectly contribute to glutamate excitotoxicity, astrocyte and microglia activation, and free radical production (Cheng et al., 2016). Neuroinflammatory cytokines in blood, serum, cerebrospinal fluid, and brain tissue have been shown to distinguish between different types of synucleinopathies (Wang et al., 2023). In patients with LBD, neurologic damage is accompanied by an inflammatory response, which may lead to increased permeability of the BBB, allowing proteins in the cerebrospinal fluid to penetrate into the bloodstream, which in turn raises Qalb levels.

Significant differences (p < 0.05) were observed when comparing Qalb levels in patients with LBD and AD, which may result from differences in BBB permeability between the two diseases. However, other mechanisms and influences may also be involved in the regulation of Qalb levels. It is suggested that Qalb may differ in the pathogenesis of AD and LBD.

AD and DLB are the most common neurodegenerative dementias, and both are characterized by β-amyloid (Aβ) misfolding and Aβ peptide deposition in the senile plaques (Aβ deposition) (Glenner et al., 1984). Musaeus et al. demonstrated that Qalb was higher in DLB patients than in AD patients (Musaeus et al., 2020). DLB is characterized by the deposition of 300 Lewy bodies, which are intraneuronal abnormal deposition of Aβ (Capouch et al., 2018). The cause of the accumulation of Lewy vesicles is largely unknown, but studies have shown that high concentrations of albumin increase fibrillation of α -synaptic nuclear proteins, which induces the formation of Lewy vesicles (Uversky et al., 2002; Munishkina et al., 2004). It has been hypothesized that albumin in the CSF of patients with DLB not only measures BBB permeability, but may be part of the mechanism of Lewy body formation. Two studies have found that human plasma albumin (HSA) significantly reduces α-syn aggregation and fibrillation, and it has even been proposed that HSA has a chaperonin-like ability to resist amyloid aggregation, which is the key to AD pathology (Kakinen et al., 1860; Bellomo et al., 2019). It is generally believed that damage to the BBB in patients and mouse models of AD accelerates the onset or progression of the disease, but the results of one study suggest that the BBB remains intact in a variety of preclinical models of AD (Bien-Ly et al., 2015). More research is needed to verify which explanation reflects the results of this study regarding Qalb in the LBD.

We found that Qalb levels were higher in patients with LBD than in patients with FTLD (p < 0.05). One possible mechanism is the difference in the pathologic process of different types of dementia diseases. LBD and FTLD differ in the characterization of neurological damage, which may lead to changes in BBB permeability and thus affect Qalb levels.

In addition, age, disease duration, and severity of disease may also be influential factors in differences in Qalb levels. Longer disease progression, higher age, and greater severity of disease may result in more pronounced BBB damage, which in turn may affect Qalb levels.

FTLD disorders are a group of clinically and pathologically distinct dementia syndromes. Frontotemporal dementia (FTD) is the second most common early-onset dementia with clinical manifestations of progressive behavioral changes, executive dysfunction, and language difficulties. The three clinical syndromes of behavioral variant FTD (bvFTD), semantic dementia (SD), and progressive non-fluent aphasia (PNFA) constitute FTLD.

Higher Qalb levels in DLB patients than in FTD were observed in a study by Skillbäck et al. (2017). This is consistent with our findings. A study of postmortem cerebral microhemorrhages in patients with FTLD detected by 7.0 T MRI found that the predominance of cortical microhemorrhages in AD and FTLD was similar to that in LBD suggesting a similar origin. They are not associated with associated cerebrovascular disease, but may be due to brain barrier disruption associated with the neurodegenerative process itself (De Reuck et al., 2012). In terms of the discriminatory potential of Qalb between patients with LBD and FTLD, we identified only two studies; therefore, further studies are needed before definitive conclusions can be made about the reliability of this biomarker.

For many years, Qalb was considered to be only an age-related parameter, but more recently it has also been linked to sex (Cosgrove et al., 2007). Exploring sex differences in the brain may be the key to understanding the role of sex in susceptibility to neurological disorders because of sex differences in brain development, structure, and neurotransmission (McCarthy et al., 2012). The estrogen 17β-estradiol may drive differential expression of enzymes involved in BBB breakdown, ultimately leading to a protective effect on the BBB (Na et al., 2015; Gu et al., 2017; Castellazzi et al., 2018; Yin et al., 2018). It is evident from literature data that males have higher Qalb values than females that seems not dependent from the underlying pathology. Our research is limited by the lack of specific Qalb data of relevant populations (including groups according to gender, age, cognitive level, onset age and course of disease). Thus, this could be a limitation of the current study.