The diagnosis of Alzheimer's disease (AD) must be restricted to people who exhibit positive biomarkers along with specific disease phenotypes (1). Behavioral and psychological symptoms are close to universally present in AD patients, and may include psychotic symptoms (PS) (2).

Psychotic symptoms are defined as the presence of delusions and/or hallucinations (2, 3), these being the most widely used diagnostic criteria for psychosis, as proposed by Jeste and Finkel (4) and revised by Cummings et al. (3). Over half of people with AD experience PS during their illness (5, 6). In particular, systematic reviews indicate a cross-sectional prevalence of approximately 40%, although it is recognized that lower rates occur in community populations and higher rates in clinical settings (2).

Psychotic symptoms have been found to be associated with rapid cognitive decline in AD (7, 8), and their emergence may represent a distinct phenotype (9). However, the association between AD psychosis and cognitive decline does not appear to be attributable to factors such as age at onset of AD, disease duration, sex, race, education, or family psychiatric history (10). PS are also associated with more rapid progression of functional impairment, hospital admission, earlier admission to institutional care, and increased mortality (11–15). Some studies have suggested that a sharper trajectory of decline occurs among people who develop PS, even before the onset of PS (8, 12, 16, 17). The use of antipsychotic drugs to treat PS in dementia is associated with greater mortality (18) and with adverse events (19).

This evidence suggests that a different underlying biological and/or genetic predisposition may be present in these individuals, and that the occurrence of PS represents a more severe AD phenotype (20); currently, however, there are no methods for predicting the occurrence of psychosis in AD patients.

The postulated neural mechanisms of AD psychosis include disturbances in cholinergic muscarinic receptors and altered concentrations of serotonin, tau protein, kalirin, and dopamine receptors (2). AD may be associated with selective alterations in dopamine receptor density (21, 22). Overexpression of wild-type alpha-synuclein (AS) decreases dopamine neurotransmission (23), and dopamine modifies aggregation of AS in the nervous system, resulting in greater abundance of AS oligomers (24). Lewy bodies, which are insoluble aggregates composed mainly of phosphorylated AS, are found in approximately 30%−50% of people with AD (25, 26); their presence contributes to the risk of psychosis and excess cognitive burden (21).

Various authors have published reports of alterations in CSF levels of AS in patients with AD, including in the prodromal phase of the illness (27–38). Increased CSF levels of AS have been found in patients with prodromal AD compared with controls and in patients with MCI attributable to Lewy body disease. Notably, however, a small subgroup of patients with MCI-AD and PS have been found to display low levels of CSF AS (37).

This study aimed to extend the sample size with a large cohort and to estimate the diagnostic validity of CSF levels of AS as a predictor of the emergence of PS in patients with prodromal AD.

This study indicated that AS is a valuable CSF biomarker for prediction of PS in a prodromal AD patient cohort. Among the other CSF biomarkers, p-tau₁₈₁, AS/p-tau₁₈₁ ratio, and Aβ₄₂ also reached the threshold, but displayed lower diagnostic validity for this purpose.

The presence of psychotic symptoms in AD is reportedly associated with AS and/or tau cerebral pathologies (8). The pathogenic relationship between AS and PS is recognized because of their common implication of cerebral dopamine levels (23, 24). Increased AS affects dopamine neurotransmission at multiple levels, particularly in decreasing dopamine synthesis (23). Moreover, dopamine influences the aggregation of AS in the nervous system, which results in AS oligomers and unique dopamine-induced oligomeric conformations (24). Selective alterations in dopamine receptor density have been found postmortem and in vivo in patients with AD and PS; these alterations may be associated with distinct clinical profiles (21, 22).

In the present study, lower CSF levels of AS were found in the AD-PS group than in the AD-No PS group. Such a decrease may be associated with the cerebral deposition of AS, forming Lewy bodies in alpha-synucleinopathies (46, 47). A number of studies have reported AS pathology in approximately 50% of autopsied patients with AD (25, 36). Patients with AD and AS tend to exhibit amplified deterioration, typically enduring more severe symptoms and shorter duration of survival (47), which may contribute to a distinctive clinical profile within AD (9, 22). Evidence from other studies suggests that AS might be involved in the development of AD from the very early stages of Aβ pathology formation (48), as well as tau hyperphosphorylation (26). These data suggest that AS is involved in the pathophysiology of AD (26). Early identification of this condition in patients with AD, including quantification of CSF AS, should enable provision of a better treatment plan and improvements in prognosis (37).

The AD-No PS group exhibited higher levels of AS than the control group, particularly during the first 4 years after diagnosis. These results were in line with those of previous studies, with short follow-up periods and without follow-up, than have taken PS into account in the clinical description or evolution (27–29, 32–36). The increase in AS observed in AD patients is based on evidence on elevated AS in the brain tissue of patients with AD (49) and/or the neuronal damage related to AD (33). This increase in CSF levels of AS in AD patients is associated with the accumulation of amyloid plaques (26) and tau proteins (36, 50, 51). These results are in accordance with the lower levels of Aβ₄₂, higher levels of p-tau₁₈₁, and higher AS/p-tau₁₈₁ ratio observed in the AD-No PS group in the present study. In a previous study, PS was found to be associated with tau phosphorylation abnormalities (8). Nevertheless, although that association should be female-specific, the incidence of PS in AD should be more likely among men because of the nature of AS pathology (8). To our knowledge, no gender differences have been described in the relationship of PS with Aβ₄₂ protein and the ratio of AS/p-tau₁₈₁.

In this study, the decrease in the aforementioned ratio among the AD-PS group was more attributable to lower levels of AS than to a clear decrease in the levels of p-tau₁₈₁. Another independent report, on a longitudinal study of AD, has found that lower values for this ratio predict faster cognitive decline (35).

The sensitivity of AS in differentiating between the AD-PS and AD-No PS groups is notable, considering the ability to predict the occurrence of PS, which requires specific pharmacologic intervention. The emergence of PS is associated with endogenous and exogenous factors, including underlying biological and/or genetic predispositions in the individual (2). Regarding endogenous factors, of the dopaminergic factors mentioned above, certain alterations to the cholinergic (2, 52–56) and serotoninergic systems (57, 58) are related to the occurrence of PS in AD. In terms of exogenous factors, sleep quality, familial relationships with caregivers, medical antecedents, schooling, and capacity to engage in activities of daily living should influence the emergence of PS in AD (2). No significant differences at the time of inclusion were found between the AD groups on any of these variables, including demographic and radiological data, indicating the high level of homogeneity between the groups included in this study.

Clinical NPI and UPDRS (III) scores increased among the PS group over the period following inclusion. These data may be concordant with abnormal dopaminergic status, which is probably related to the amount of cortical and subcortical Lewy bodies (59). To date, the concept of mixed AD + dementia with Lewy bodies is accepted in neuropathological settings as a difficult clinical diagnosis, with a lack of biomarkers to assist identification (60). Quantification of CSF level of AS may facilitate this objective (61). Positive correlations were observed between AS levels, such scores, and tau protein levels in both AD groups and in the subgroups. These findings in the initial stages of AD support the involvement of AS in the pathogenesis of AD (27, 32, 52).

Alzheimer's disease is considered to be a clinical–biological entity (1). Our data and data presented in previous publications (27, 28, 32–37) support the added value of measurement of CSF levels of AS in further characterization of the CSF AD biomarker profile. The contradictory results published in some previous reports regarding the potential use of CSF level of AS as a diagnostic biomarker for AD may be attributable to various factors, including reduced sample size, uncertain diagnosis and/or uncontrolled follow-up, enrolment of patients at different stages of the disease, differences in age, and lack of control for blood contamination, among others (29, 62–65). The current AT (N) classification for a biological definition of AD has been defined as flexible, meaning that new biomarkers can be added when they become available (40), and there is a need for more biomarkers, such as AS, indicating other aspects of the mechanisms of the disease (66–69).

Beta-synuclein is another member of the synuclein family and is emerging as a reliable synaptic marker in CSF and blood for AD and prion disease (70, 71). In the study by Barba et al., increased levels of CSF AS were observed in pre-AD patients, but not in MCI-AD or dementia-AD. Beta-synuclein was elevated in all AD continuum subgroups. Elevated CSF levels of both beta-synuclein and AS in pre-AD may reflect the earliest synaptic dysfunction occurring in AD. Decreased AS levels may indicate the presence of α-synucleinopathy, whereas beta-synuclein concentrations are not influenced by the presence of synucleinopathy or by blood contamination, which instead affects AS measurements.

In this study, the handling of the samples was performed as per recommended operating procedures (72). CSF samples with < 50 red blood cells/μL were included (28), and the reagents used had been validated in a Europe-wide inter-laboratory study (45). These points of analytical and methodological validity are critical for the value of these results. Nevertheless, in relation to protein-misfolding cyclic amplification (PMCA) and real-time quaking-induced conversion (RT-QuIC), ultrasensitive protein amplification assays for the detection of misfolded protein aggregates could also offer diagnostic reliability for AS levels in CSF. In particular, RT-QuIC assay has demonstrated high specificity and sensitivity for detection of CSF AS aggregation in patients with synucleinopathies when compared to AD patients and controls (73–75). The possibility of analyzing the cholinergic, glutamatergic, and serotoninergic pathways with these techniques should be of great interest for completion of the study of PS in AD patients.

The main limitation of this study was the lack of neuropathological confirmation. Nevertheless, to our knowledge, the long clinical follow-up period of the patients was among the longest across all published studies. The clinical and biological data were very conclusive. However, the ability of caregivers to control behavioral disturbances was not tested, and a portion of the AD-No PS group had an incomplete period of follow-up in which the emergence of PS could be excluded. Finally, the lack of data on APOE genotype in patients recruited during the first 5 years of the study was another limitation.

In conclusion, the quantification of CSF AS in patients with prodromal AD allows for prediction of the emergence of PS during the subsequent 8 years with high sensitivity. In this regard, early identification will enable the provision of better treatment plans and improvements in prognosis. Future biomarker panels, including the biological study of cholinergic and serotoninergic pathways, will probably enable more complete prediction of PS in AD.

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

The studies involving human participants were reviewed and approved by Ethical Committee of University General Hospital Dr. Balmis. Alicante, Ref number: PI2020/250. The patients/participants provided their written informed consent to participate in this study.

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.