Plasma Proteins, Sex and Alzheimer’s Disease: Proteome-Wide Analyses of the UK Biobank and Framingham Heart Study

Nearly two-thirds of Americans with Alzheimer’s disease (AD) are women. While this higher prevalence is due in part to women living longer than men, emerging research indicates that sex-related differences in multiple factors—genetics, hormonal changes and other lifestyle factors—can also contribute to this disparity. Understanding which factors are major contributors to sex differences in the onset and progression of AD is expected to help guide the development of earlier diagnosis and more effective prevention and treatment strategies for both sexes.

To address some foundational questions about the factors that impact sex differences in AD, Dr. Doraiswamy and his team turned to the long-running Framingham Heart Study (FHS). They focused on the FHS “Off-Spring” cohort, for which extensive demographic, medical and other clinical data were collected for up to 13 years as well as measurements of 30 blood-based protein biomarkers for each participant. They used this rich dataset to explore the associations between sex, AD risk and several factors of interest (age, APOE status, menopause, biomarkers). Their results confirmed that women in this cohort were more likely to be diagnosed with AD, but they also found this was primarily true for women aged 60-70 and in those with lowest levels of education (no high school). They also reported that menopause at an earlier age—often due to surgery—increased AD risk. For the blood-based biomarkers, they found several intriguing, yet still preliminary, changes in protein levels that predicted the rate of cognitive decline better in women than men. Despite these new insights, combining demographic factors, cognitive performance, ApoE4 status, and this subset of blood biomarkers accounted for only ~20% of the total AD risk for both sexes, leaving many aspects of risk still unexplained.

Here, Dr. Doraiswamy proposes to follow up and significantly expand on the blood-based biomarker results. To increase the number of samples and proteins measured, the team will leverage two of the most comprehensive cohort studies available: the FHS again and the UK Biobank. The UK Biobank is comprised of samples and data from approximately 500,000 middle-aged and older participants. It includes clinical data as well as multiple other types of biological data (e.g., genetics, brain imaging). Notably, the UK Biobank recently made new ‘proteomic’ data available to researchers—measurements of 3,000 different proteins, detected in blood samples from over 50,000 people. Likewise, the FHS now has similar proteomic data for these 3,000 proteins in 5,800 people. The team predicts that identifying sex-specific changes in blood biomarkers that are associated with a later AD diagnosis will enhance the ability of clinicians to predict who is at risk. Their goal in this project is to discover these changes in midlife—at least a decade prior to the onset of AD.

They proposed four aims. In the first aim, they will confirm whether there are similar sex differences in clinical AD incidence within the UK Biobank cohort. They expect that sex, education level, plasma amyloid and ApoE4 status will impact AD risk in this cohort like they reported for the FHS. In the second aim, they will use the available blood protein biomarker data to first discover any that are associated with an AD or dementia diagnosis in men and women from the UK Biobank. Then, they will see if these data are replicated in the FHS cohort. Special attention will be given to established biomarkers (e.g., pTau181) to assess whether they exhibit sex-specific variations. In the third aim, the team will use the same biomarker data to find proteins that change in relation to neurodegeneration, visible via MRI scans. These findings will also be validated in the FHS cohort. For their fourth aim, which they call exploratory, they plan to develop a sex-specific ‘score’ based on blood biomarker data that reliably predicts a future diagnosis of AD.

They made strong progress in the first year of funding. Using data from the UK Biobank and replication in the Framingham Heart Study, the team examined how genetic risk, including APOE genotype and Alzheimer’s polygenic risk scores, influences blood-based protein signatures associated with future AD diagnosis. Their analyses identified genetic risk–dependent proteomic pathways and revealed both shared and sex-specific biomarkers linked to incident AD, including replication of established markers such as GFAP and NEFL. These findings suggest that incorporating sex-specific biomarker profiles may improve prediction of future AD risk and provide new insight into early biological changes that precede cognitive decline. In the next year of funding, the team plans to build on these findings by investigating how sex-specific blood-based biomarkers relate to MRI measures of neurodegeneration and structural brain changes. Using longitudinal biomarker and imaging data, they aim to further define early disease mechanisms and improve approaches for identifying individuals at risk for AD before symptoms emerge.

Together, these findings could lead to important insights into early biological changes that precede cognitive decline and provide new opportunities to detect at risk individuals and intervene early.