AD-Detect™ Likelihood Score

Alzheimer's disease quietly builds in the brain for one to two decades before the first noticeable memory slip. The biology that defines it, sticky protein deposits called amyloid plaques, can now be hinted at through a blood draw rather than a brain scan or spinal tap. The AD-Detect Likelihood Score is one of the first commercial tools designed to do this from a single tube of blood.

For anyone with a family history of dementia, early memory concerns, or a personal interest in protecting cognitive health, this score offers a window into a process that standard cognitive testing cannot see. It will not diagnose Alzheimer's by itself, but it can flag whether the biology of the disease appears to be present, raising or lowering the priority of further workup.

What the Score Actually Reflects

The AD-Detect Likelihood Score is built on the ratio of two protein fragments measured in blood plasma, Aβ42 and Aβ40 (two forms of the amyloid beta protein that gets stuck in Alzheimer plaques). When the brain is accumulating plaques, the relative amount of Aβ42 in the bloodstream tends to drop. The test measures these two fragments using a sensitive lab method, then converts the ratio into a likelihood category.

A higher likelihood score means the blood pattern looks more like that of people who have brain amyloid on imaging. A lower score means the pattern looks more like that of people whose brains are clear. The score is a probability statement about brain biology, not a diagnosis of dementia.

Alignment With Brain Amyloid Imaging

The most rigorous test of any blood-based Alzheimer's marker is how well it matches a brain scan that directly visualizes amyloid plaques. In validation work using the same plasma Aβ42/40 assay that underlies AD-Detect, the ratio identified most amyloid-positive scans and correctly cleared most amyloid-negative scans, with sensitivity in the roughly 77 to 88 percent range and specificity in the roughly 73 to 78 percent range across cohorts. That is good but not at the level of brain imaging on its own.

Plasma scores that combine the amyloid ratio with phosphorylated tau (a related Alzheimer's marker) and other features perform better. In large primary and specialty care studies, blood scores built from these components agreed with amyloid imaging the large majority of the time. Recent guidelines accept high-performing blood tests as a substitute for a brain scan in patients being evaluated for memory loss, but those guidelines set a roughly 90 percent sensitivity and specificity threshold that is generally reached by phosphorylated tau 217-based tests rather than by the amyloid ratio on its own.

Risk of Cognitive Decline and Dementia

A low Aβ42/40 ratio, which produces a higher AD-Detect Likelihood Score, has been tied to faster cognitive slippage and a higher chance of progressing to dementia. In adults with mild cognitive impairment (the in-between state where memory is worse than expected for age but daily life is preserved), those with abnormal plasma amyloid made far more semantic intrusion errors (a specific kind of memory mistake that is sensitive to early Alzheimer's) than those with normal plasma amyloid.

In a community study of more than 2,000 dementia-free older adults, elevated AD blood biomarkers predicted dementia within 10 years with a strong ability to rule out future disease in those who tested normal (a negative predictive value above 90 percent). The flip side is that a positive result in a healthy person predicts disease less reliably, which is why the score is best understood as one input into a broader picture, not a verdict.

Distinguishing Alzheimer's From Other Causes of Memory Loss

Memory problems can stem from many things, including vascular disease, frontotemporal dementia, Lewy body disease, sleep disorders, depression, thyroid disease, and medication side effects. Most of these do not produce a low plasma Aβ42/40 ratio, though some conditions involving brain amyloid (such as cerebral amyloid angiopathy or Lewy body disease with co-occurring amyloid) can. That makes the AD-Detect score most useful when the question is specifically whether Alzheimer's biology is contributing to symptoms. A high likelihood score points toward amyloid as part of the picture. A low likelihood score makes Alzheimer's a less likely explanation and pushes the workup toward other causes.

Why One Reading Is Not Enough

Plasma amyloid markers show meaningful short-term variability between blood draws in the same person. Studies in memory clinic patients found that within-person fluctuation can be large enough to shift a borderline result into a different category from one visit to the next. The most reliable interpretation comes from confirming an abnormal result with a repeat test rather than acting on a single value.

For someone with an intermediate or high-likelihood result, a confirmatory blood draw before any decision about further imaging or specialist referral is a sensible step used in research and specialty practice, although exact timing is not set by published guidelines. Routine serial screening in people without cognitive symptoms is not currently endorsed by major guideline bodies such as the Alzheimer's Association or the American Academy of Family Physicians. The goal of any repeat draw is to see the trajectory rather than to act on a single number.

When Results Can Be Misleading

A few factors can distort the picture and should be considered before acting on a result:

• Reduced kidney function: plasma levels of amyloid and tau fragments rise when the kidneys clear them less efficiently. The Aβ42/40 ratio is more robust to this effect than individual analytes because both numerator and denominator rise together, but significant chronic kidney disease still warrants cautious interpretation.
• Recent stroke or major medical illness: acute brain injury and serious systemic illness can temporarily move blood-based brain markers, and real-world studies have shown reduced diagnostic reliability after stroke and in patients with high overall comorbidity.
• Borderline values: results that fall near the cutoff between low and high likelihood deserve a confirmatory draw rather than action, since short-term biological variability can move a borderline score across the line.
• Anti-amyloid antibody therapy: if a person is being treated with a drug that clears brain amyloid, the score will shift in response to treatment and should not be interpreted as a baseline risk reading during therapy.

What to Do With an Out-of-Pattern Result

A high-likelihood result is a prompt to investigate, not to panic. A reasonable approach used in specialty practice includes a confirmatory blood draw, followed by either a phosphorylated tau 217 (a more specific blood marker of Alzheimer's tau pathology) measurement, a brain amyloid scan, or cerebrospinal fluid (the liquid surrounding the brain and spinal cord) testing to confirm the biology. Pairing the result with a structured cognitive assessment and, where relevant, an APOE (apolipoprotein E, a gene that influences Alzheimer risk) genotype helps stratify how aggressive the workup should be. This stepwise sequence is a clinical synthesis, not a published algorithm.

If memory or thinking symptoms are present, a referral to a neurologist or memory specialist is appropriate, and this is the setting in which current guidelines most strongly support blood-based biomarker use. If the result is high but cognition is intact, the action shifts toward modifiable risk factors (sleep, blood pressure, glucose control, exercise, hearing) rather than immediate treatment, since major guideline bodies do not currently recommend blood-based Alzheimer's testing or treatment in people without cognitive symptoms. A low-likelihood result with normal cognition is reassuring but does not rule out future amyloid accumulation.