Decoding subphenotypes in electronic medical records within late-onset Alzheimer's disease reveals heterogeneity and sex-specific differences

medRxiv [Preprint]. 2025 Apr 25:2025.04.24.25326306. doi: 10.1101/2025.04.24.25326306.

ABSTRACT

BACKGROUND: Alzheimer's disease is a progressive neurodegenerative disorder with no curative treatment. Identifying distinct subphenotypes and understanding potential personalized modifications remain critical unmet needs.

METHODS: We applied unsupervised learning techniques to electronic medical records from UCSF to identify distinct Alzheimer's disease subphenotypes based on comorbidity profiles. We conducted enrichment analyses to determine cluster-specific comorbidities. Based on the observed sex-based differences, we subsequently conducted sex-stratified analyses to assess differences in disease manifestations between males and females. Findings were validated using an independent UC-Wide dataset.

RESULTS: Among 8,363 patients, we identified five Alzheimer's disease subphenotypes, characterized by comorbidities related to cardiovascular conditions, gastrointestinal disorders, and frailty-related conditions such as pneumonia and pressure ulcers. Sex-stratified analyses revealed significant differences in comorbidity distributions across clusters. Notably, in Cluster 2, circulatory diseases were more prevalent among males, whereas in Cluster 3, bladder stones were more common among females. Key results were consistent across the UCSF and UC-Wide datasets.

CONCLUSIONS: Our study identifies clinically meaningful Alzheimer's disease subphenotypes and highlights sex-specific variations, suggesting potential underlying biological factors such as Apolipoprotein E and gut microbiome alterations contributing to Alzheimer's disease heterogeneity. These findings underscore the need for further research into the biological mechanisms driving these differences and may inform the development of individualized therapeutic regimens.

FUNDING: This study was supported by grants from the National Institute on Aging (R01AG060393 and R01AG057683).

PMID:40313288 | PMC:PMC12045436 | DOI:10.1101/2025.04.24.25326306