New Alzheimer’s Diagnostic: Blood Flow vs. Amyloid

Impaired Brain Blood Flow Might Be Important in Alzheimer’s

A CO2-based diagnostic was far more accurate than an amyloid-based one.

by

August 26, 2025

Introduction

Under the vascular hypothesis of Alzheimer’s, the disease is primarily characterized by the blood that the brain receives. A diagnostic test based on this hypothesis outperformed other, well-established diagnostics.

Scientists have developed a potent diagnostic tool based on the vascular hypothesis for Alzheimer’s, which might offer clues to the mechanism behind the disease [1].

The Blood Flow Hypothesis

Traditionally, Alzheimer’s disease has been associated with the presence of amyloid plaques and tau tangles in the brain, known as the amyloid cascade hypothesis. However, recent studies have raised doubts about this model [2]. Notably, researchers have observed that many individuals with significant amyloid buildup never develop dementia [3], and that cerebrovascular disease is a common feature of Alzheimer’s [4]. This has led to the development of the vascular hypothesis, which posits that problems with the brain’s blood supply are a key factor in the disease.

In a study from the University of Southern California, scientists propose that a breakdown in cerebral perfusion dynamics leads to reduced blood flow (hypoperfusion) and hypoxia. The onset of hypoxia can increase the production of amyloid, resulting in a vicious cycle where vascular damage and amyloid pathology reinforce each other and ultimately lead to cognitive decline.

Building the New Diagnostic Tool

Existing early diagnostics involve either a spinal tap or a PET scan to measure amyloid content. “Physicians take the emissions from that PET radioactive tracer as an approximate measure of how much amyloid or tau the person has in their brain,” said Vasilis Marmarelis, Dean’s Professor in the Alfred E. Mann Department of Biomedical Engineering. “Speaking from experience, after having seen data in my own study, I can tell you that it’s very inadequate. But it’s the gold standard, although most physicians don’t do it because it’s very expensive.”

Another popular option involves cognitive assessment tests, such as the Mini-Mental State Examination (MMSE) and the Montreal Cognitive Assessment (MoCA). However, they also have significant limitations. “The MoCA does give you an idea about the cognitive abilities of a person,” Marmarelis commented. “But these are not biological or physiological tests. These are just behavioral tests that are subject to various biases, imperfections, and errors.”

The researchers recruited 167 participants and categorized them into three groups: cognitively normal, mild cognitive impairment (MCI), or mild Alzheimer’s disease. Each participant underwent a non-invasive procedure where researchers recorded several signals simultaneously. This included:

• Transcranial Doppler to measure blood flow speed in major brain arteries.
• Near-infrared spectroscopy to measure oxygen levels in the prefrontal cortex.
• Continuous monitoring of arterial blood pressure and end-tidal CO₂.

“When we exert cognitive effort, we generate CO₂ from the metabolism in our cerebral cells, which has to be removed by our blood to prevent acidosis,” Marmarelis explained. “Our body is endowed with a regulatory mechanism called vasomotor reactivity, which dilates cerebral vessels when CO₂ levels rise in the blood, ensuring adequate blood flow and removal of excess CO₂.”

CDI Beats the Rest

Given the importance of early diagnostics of Alzheimer’s, the team aimed to develop and test a novel marker called the Cerebrovascular Dynamics Index (CDI). This is a comprehensive score that assesses:

1. How well blood vessels adapt to changes in CO₂.
2. The brain's ability to maintain stable blood flow during blood pressure fluctuations.
3. Oxygen levels in the cortex during variations in both blood pressure and CO₂.

CDI exhibited exemplary performance compared to conventional methods used as positive controls. The study measured performance using “area under the curve” (AUC), where 1.0 indicates a perfect test and 0.5 is no better than chance.

CDI also excelled in determining disease severity, distinguishing MCI from Alzheimer’s. It achieved an AUC of 0.98, surpassing MMSE (0.83) and MoCA (0.78), while outmatching the amyloid PET scan (0.61), which was not even statistically significant, suggesting that amyloid levels alone were inadequate for this purpose.

“What we have that others didn’t have before is a methodology to quantify these dynamic relations that’s extremely robust and accurate,” Marmarelis stated. “We can now differentiate patients with mild cognitive impairment and Alzheimer’s from cognitively normal controls far better than the PET measurement and even better than the MoCA neurocognitive test.”

Implications for Future Research

This study's significance extends beyond diagnostics; it may reveal latent truths about Alzheimer’s and influence therapy development. According to Marmarelis, “the particular aspect of dysregulation of cerebral perfusion regulation may be the critical factor in the pathogenesis of this disease, likely in conjunction with other factors, including amyloid accumulation.”

For those interested in supporting independent journalism, we, as a non-profit foundation, offer all our news and educational content for free. Every contribution, regardless of size, sustains our future and the mission of advancing knowledge and understanding.

Literature Cited

[1] Marmarelis, V., et al. (2025). Dysregulation of cerebral perfusion dynamics is associated with Alzheimer’s disease. Alzheimer’s & Dementia: Diagnosis, Assessment & Disease Monitoring, 17(3), e70134.

[2] Kepp, K. P., et al. (2023). The amyloid cascade hypothesis: an updated critical review. Brain, 146(10), 3969-3990.

[3] Aizenstein, H. J., et al. (2008). Frequent amyloid deposition without significant cognitive impairment among the elderly. Archives of Neurology, 65(11), 1509-1517.

[4] De la Torre, J. C. (1994). Impaired brain microcirculation may trigger Alzheimer’s disease. Neuroscience & Biobehavioral Reviews, 18(3), 397-401.