Mild Cognitive Impairment (MCI) is a transitional state between normal aging and dementia, frequently preceding Alzheimer's disease (AD). Identifying MCI individuals at high risk of converting to dementia is crucial. While the role of genetic factors like APOE ε4 and neuropsychiatric symptoms (NPS) in dementia risk is established, their combined effect, particularly in a clinical setting, remains understudied. Only a few studies have explored the interaction between NPS and APOE ε4 in predicting dementia risk, primarily in population-based cohorts rather than memory clinics. This study aims to determine the interaction effect of NPS and APOE ε4 on dementia conversion in a large sample of MCI patients from a memory clinic, controlling for various clinical factors.

Previous research has highlighted the individual contributions of both APOE ε4 and neuropsychiatric symptoms (NPS) to the risk of developing dementia and Alzheimer's disease. Studies have shown that individuals carrying the APOE ε4 allele have a significantly increased risk of developing AD, and the presence of NPS is often associated with faster cognitive decline and increased dementia risk. However, few studies have specifically investigated the interaction between these two risk factors, and even fewer have done so within the controlled setting of a memory clinic. Existing research, often population-based, suggests additive or even synergistic interactions between specific NPS and APOE ε4 in predicting dementia risk. This study aims to address the gap in knowledge by examining these interactions within a memory clinic population, offering a more clinically relevant context.

This prospective cohort study included 1512 MCI patients (aged 60+) from the Fundació ACE Memory Clinic (Barcelona, Spain) followed from 2005 to 2018. Participants met Petersen's MCI criteria. Neuropsychiatric symptoms were assessed at baseline using the Neuropsychiatric Inventory-Questionnaire (NPI-Q), which evaluates 12 symptoms including depression, apathy, anxiety, agitation, etc. Neuropsychological testing (NBACE) was conducted to assess eight cognitive domains. APOE genotyping was performed using TaqMan assays. Patients were classified as MCI converters (developed dementia) or non-converters based on Clinical Dementia Rating Scale (CDR) scores. Cox proportional hazards models were used to analyze the association between NPS, APOE ε4 status, and dementia conversion, exploring both additive and multiplicative interactions. Models were adjusted for age, sex, education, baseline MMSE, memory pattern, family history of neurological conditions, medical comorbidities, and antidepressant/anxiolytic medication use. Statistical significance was set at p<0.05.

At baseline, the prevalence of NPS varied widely (55% for depression and anxiety to <0.5% for elation/euphoria). 31% of the sample were APOE ε4 carriers. The mean follow-up was 2.03 years. 67.9% of patients were amnestic MCI. Dementia conversion occurred in 58% of patients. Older age, female sex, lower education, APOE ε4 carrier status, lower baseline MMSE scores were more common among converters. Significant additive interactions between APOE ε4 and several NPS (depression, apathy, anxiety, agitation, appetite, irritability) were observed in predicting dementia conversion, even after adjusting for multiple clinical variables. These interactions significantly increased the hazard ratios for dementia conversion (HR range 1.3–2.03). Agitation/aggression and appetite showed the highest incremental dementia risk (>85%) when combined with APOE ε4. Apathy was the only NPS exhibiting a significant multiplicative interaction, but with an inverse effect (HR=0.74). In analyses differentiating between Alzheimer's disease and other dementias, apathy was significantly associated with non-AD dementia conversion.

This study provides the first evidence of interactions between NPS and APOE ε4 in predicting dementia conversion in MCI patients within a memory clinic setting. The findings confirm the importance of considering the combined effect of these risk factors, going beyond analyses of their individual contributions. The synergistic effect observed between specific NPS and APOE ε4 highlights the potential for a combined detrimental impact on dementia risk. These interactions, particularly the strong effects for agitation/aggression and appetite, are clinically significant, providing markers for identifying patients at heightened risk of dementia. The results are largely consistent with previous population-based studies, indicating generalizability across settings. The absence of strong multiplicative interactions, except for the inverse effect of apathy, suggests additive mechanisms as primary drivers of the combined effect. The robust findings even after adjusting for various clinical variables suggests a strong and consistent association between the combination of these factors and dementia risk.

This study demonstrates additive interactions between several NPS (depression, apathy, anxiety, agitation, appetite, irritability) and APOE ε4 in predicting dementia conversion in MCI patients. The convergence of these factors significantly increases dementia risk beyond the individual effects of each. These findings offer a valuable tool for identifying high-risk MCI individuals in clinical practice, enabling early intervention strategies to mitigate the neuropsychiatric symptoms and potentially delay or modify the course of dementia. Future studies should focus on validating these findings with longer follow-up periods and explore the underlying biological mechanisms driving these interactions.

The study has some limitations. The relatively short follow-up period (2 years) might underestimate the long-term impact of these interactions. The exclusion of elation/euphoria and motor disturbances due to low prevalence might have biased the results. The analysis included all forms of dementia, not only Alzheimer's disease, which could have diluted the effect of APOE ε4. Further research with longer follow-up, larger samples, and a focus on specific dementia subtypes is needed to fully elucidate these relationships.