Longitudinal Blood Pressure Fails to Track With Composite Cerebral Amyloid Angiopathy
In a large autopsy-based analysis of older adults, longitudinal blood pressure did not align with moderate-to-severe cerebral amyloid angiopathy in the composite neocortex. The finding narrows a question that has lingered in aging and dementia research: whether repeated blood pressure measurements across later life meaningfully track the vascular amyloid burden that accumulates in cerebral vessels as people approach the end of life.
The cohort was substantial. Investigators analyzed 2,107 older adults with a mean age of 89.63 years, drawn from five longitudinal studies of aging. Blood pressure variables included systolic blood pressure, diastolic blood pressure, mean arterial pressure, and pulse pressure, each measured at baseline and during annual clinical visits. Cerebral amyloid angiopathy and Alzheimer’s disease pathology were assessed at autopsy, allowing the study to connect repeated ante-mortem hemodynamic data with postmortem neuropathology in the same participants.
That design matters because cerebral amyloid angiopathy is not a static lesion. It reflects progressive amyloid deposition in cortical and leptomeningeal vessels, a process that may respond differently to chronic perfusion pressure, age-related vascular stiffening, and concurrent Alzheimer’s pathology. In principle, higher arterial pressure could worsen vessel wall injury or accelerate amyloid-related vascular damage. The present analysis did not support that broad expectation in the composite neocortex.
Pressure Trends Across Time Did Not Explain Global CAA Burden
The primary analysis used logistic regression to test whether mean blood pressure across follow-up was associated with moderate-to-severe CAA in the composite neocortex. After adjustment for age at death, sex, and education, the association was absent. The models also incorporated cohort as well as established CAA risk factors, including APOE genotype and the presence of Alzheimer’s disease pathology.
The negative result is clinically meaningful because it suggests that average blood pressure across later life may not be a sufficient proxy for the global burden of vascular amyloid in the cortex. Blood pressure is often treated as a broad vascular risk marker, but cerebral amyloid angiopathy may be driven by more localized processes, including region-specific vessel vulnerability, amyloid clearance failure, and overlap with Alzheimer-type neurodegeneration. A single vascular summary measure may blur those distinct biological pathways.
The absence of a composite association also argues against overinterpreting blood pressure control as a direct surrogate for CAA burden in older adults. It may still matter for stroke prevention, cardiovascular risk reduction, and overall brain health, but this analysis did not show a simple monotonic relationship between mean pressure and the presence of more severe amyloid angiopathy in the neocortex as a whole.
A Small Frontal Signal Emerged in Pulse Pressure
One secondary analysis did yield a different pattern. Mean pulse pressure showed a weak inverse association with moderate-to-severe CAA in the frontal neocortex, with an odds ratio of 0.991 and a 95% confidence interval of 0.983 to 0.998. The association persisted after adjustment for Alzheimer’s disease and APOE genotype.
Pulse pressure often rises with arterial stiffening in advanced age, and it is frequently interpreted as a marker of reduced vascular compliance. A weak inverse association in one cortical region is not the same as a biologically definitive signal, but it does suggest that the relationship between hemodynamics and vascular amyloid may vary by brain region rather than follow one uniform cortical pattern. The frontal neocortex may differ from temporal, parietal, and occipital regions in vascular architecture, perfusion demands, or susceptibility to amyloid deposition.
The regional nature of the signal should temper any broad conclusion. No comparable association emerged in the composite neocortex, and the study’s regional result was modest in magnitude. Still, the frontal finding points to a question that deserves more precise mapping: whether specific hemodynamic patterns correspond to selective vulnerability in particular cortical territories.
What the Autopsy Data Can and Cannot Resolve
This analysis has a strength that many clinical studies lack: direct neuropathologic assessment. Because CAA was measured at autopsy rather than inferred from imaging or symptoms alone, the investigators could evaluate actual postmortem vascular amyloid burden. Longitudinal blood pressure collection also reduced reliance on a single clinic reading, which is especially important in very old adults whose pressure values fluctuate across visits and across stages of frailty.
Even so, the study asks a limited question and answers it within a specific setting. It is a secondary analysis of prospectively collected data, not a trial of blood pressure treatment or a mechanistic experiment. The participants were exceptionally old, with a mean age near 90, so the findings may not transfer directly to younger older-adult populations or to individuals earlier in the course of cerebrovascular disease. The analysis also focused on mean values over time rather than on variability, duration of hypertension, antihypertensive treatment patterns, or blood pressure trajectories around critical illness.
Those omissions leave room for alternative explanations. Cerebral amyloid angiopathy may depend more heavily on cumulative lifetime exposure than on late-life averages. It may also relate more to blood pressure fluctuations, vascular stiffness, or treatment intensity than to mean pressure alone. Autopsy-based studies can identify final pathology with precision, but they cannot reconstruct every biological step that led there.
Why Region-Specific Biology May Matter More Than a Global Average
The most useful takeaway from this study is not that blood pressure is irrelevant to brain vascular disease. Rather, it is that a single summary measure of blood pressure may be too crude to capture the complexity of CAA. The frontal neocortex result, although weak, hints that regional anatomy could shape how vascular pressure intersects with amyloid deposition. Small-vessel integrity, local amyloid clearance, and regional patterns of neurodegeneration may all modify the relationship.
For clinicians and researchers, that means future work will need more granular exposure measures and more granular neuropathologic endpoints. Specific blood pressure values, visit-to-visit variation, treatment exposure, and perhaps nocturnal or ambulatory profiles may be more informative than annual office measures. Regional pathology scoring may also prove more revealing than a composite cortical endpoint when the biology is heterogeneous.
The current data do not justify a claim that higher mean blood pressure drives composite neocortical cerebral amyloid angiopathy in the very old. They do justify a narrower, more technical question: whether selected pressure features, particularly pulse pressure, map onto regional CAA severity in ways that standard averages conceal. The study stops there, and the unresolved issue is whether those regional patterns hold in other cohorts, other age ranges, or with more detailed blood pressure phenotyping.