Neuropsychiatric symptoms (NPSs), encompassing behavioral and psychological changes like mood, anxiety, and drive impairment, are linked to poorer outcomes in cognitive function, quality of life, and dementia progression. They're considered early manifestations in Alzheimer's Disease (AD), predicting a higher risk of conversion to dementia, even in preclinical and prodromal Mild Cognitive Impairment (MCI) stages. In vivo biomarkers quantifying AD-related brain changes (Aβ amyloid plaques, tau tangles, neurodegeneration, vascular changes) are now available through various methods: cerebrospinal fluid (CSF), plasma, imaging (MRI, PET), and genomics. The National Institute on Aging and Alzheimer's Association (NIA-AA) research criteria for AD now incorporate these biomarkers into the AT(N) framework, classifying individuals based on Aβ plaques (A), pathological tau (T), and neurodegeneration (N). This allows biological classification of AD regardless of clinical presentation. This study addresses the gap in understanding how AD biomarkers (AT(N) profiles) and NPSs interact to predict dementia conversion in MCI patients. It aims to determine the predictive value of combining AT(N) profiles (using CSF biomarkers) and NPSs (measured by the Neuropsychiatric Inventory Questionnaire, NPI-Q) in a memory clinic cohort followed for two years. The study investigates both additive and multiplicative interaction effects of these factors on dementia conversion rates. This is important to identify individuals at the highest risk, potentially enabling earlier interventions and personalized care strategies.

Existing research highlights the association between NPSs and poor outcomes in MCI and AD. Studies have shown that NPSs predict faster cognitive decline and increased risk of dementia conversion. However, few studies examined the relationship between AD biomarkers and NPSs, particularly their interaction effect on dementia conversion in MCI patients. While in vivo biomarkers like CSF Aβ, tau, and neurodegeneration markers are crucial in defining AD pathology, their relationship with NPSs remains an area of ongoing investigation. The AT(N) framework, though valuable in research settings, hasn't extensively explored its synergistic effect with NPSs in predicting dementia. Therefore, understanding this interaction is key to refining our prediction models and developing targeted interventions. Some research indicated a link between specific NPSs such as depression and apathy, and cerebrovascular disease, raising questions on the role of vascular dysfunction in the interplay between NPSs and AD pathology.

This prospective cohort study included 500 MCI patients from the Ace Alzheimer Center Barcelona memory clinic (2016-2022). Within five months of enrollment, patients underwent lumbar puncture (LP) for CSF biomarker analysis (Aβ1-42, total tau, p181-tau) and NPS assessment using the NPI-Q. The AT(N) classification was determined using in-house cutoffs from the ACE Alzheimer Center Barcelona CSF program, categorizing patients into normal, brain amyloidosis, prodromal AD, and SNAP groups. The NPI-Q assessed various NPSs, focusing on depression, anxiety, apathy, irritability/lability, and nighttime behaviors. Patients were classified based on the presence or absence of each NPS. MCI subtypes (amnestic/non-amnestic, possible/probable) were also determined using established criteria. APOE ε4 carrier status was ascertained through genomic DNA analysis. Annual follow-up assessments were conducted to determine conversion to dementia. Dementia subtypes were classified using standard criteria for AD, vascular dementia, frontotemporal dementia, and Lewy body dementia. Cox proportional hazard models were employed to analyze the impact of AT(N) profiles and NPSs on dementia conversion. Age, sex, education, baseline MMSE score, APOE ε4 status, and MCI subtypes served as covariates. Additive and multiplicative interaction effects between AT(N) profiles and NPSs were explored using appropriate statistical methods, focusing on comparisons between the normal and prodromal AD groups for interactions. Cumulative hazard functions were plotted to visualize the combined effects.

During the two-year follow-up, 224 participants (44.8%) converted to dementia. The prodromal AD AT(N) profile was prevalent in MCI converters (60% vs. 26% in non-converters). All pathological AT(N) groups (prodromal AD, brain amyloidosis, SNAP) showed increased risk of dementia conversion compared to the normal group (HRs ranged from 1.78 to 4.34). Among NPSs, depression and apathy predicted higher dementia risk (HRs around 1.46 and 1.40, respectively). A significant additive interaction was observed between AT(N) profiles and depression (p=0.037), indicating that depression further increased dementia risk in individuals with a prodromal AD profile. Initially, a multiplicative interaction was seen with nighttime behaviors; however, this interaction lost significance when other NPSs were included as covariates. The additive interaction between AT(N) profile and depression remained significant after adjusting for other NPSs (p=0.044).

This study demonstrates that both pathological AT(N) profiles and certain NPSs (depression and apathy) are significant risk factors for dementia conversion in MCI. The most striking finding is the synergistic effect of prodromal AD in CSF and depressive symptoms, significantly increasing the risk of dementia conversion beyond the sum of individual risks. The lack of interaction for other NPSs suggests that the additive effect of depression is specific and not merely a consequence of comorbid NPSs. The results align with previous studies suggesting depression might accelerate cognitive decline. The inconsistent results in prior studies on the relationship between AD biomarkers and NPSs are likely due to methodological differences. This study contributes a novel perspective by explicitly examining the modulating effect of AT(N) profiles on NPSs and their predictive value in dementia conversion, particularly the synergistic interaction between prodromal AD and depression.

This study provides the first evidence of a synergistic effect between CSF AT(N) profiles and depression in predicting dementia conversion in MCI. The combination of a prodromal AD profile and depression significantly exacerbates dementia risk. This highlights the importance of considering both biological and behavioral factors when assessing prognosis and developing targeted interventions. Future studies with longer follow-up periods and broader NPS assessment are necessary to confirm these findings and explore the potential for early interventions.

This study has limitations: the two-year follow-up period may be too short to fully capture the long-term impact of these factors. Several NPSs with low prevalence were excluded from the analysis, potentially affecting the generalizability of the findings. The AT(N) profile and neuropsychiatric status are not static and future studies might use more complex analytical approaches to better account for longitudinal changes.