Not every Alzheimer's patient loses their memory first. Some lose the ability to read a clock face, or find the right word mid-sentence, or plan a simple errand, months or years before their memory shows any obvious decline. A study published August 5, 2026, in the journal Neurology found that these patients, whose disease shows up atypically rather than through the classic memory-loss pattern most people associate with Alzheimer's, are being left behind by the very treatments designed to slow the disease down.

A disease that doesn't always start where doctors expect

The study, led by Dr. Dror Shir of the Mayo Clinic in Jacksonville, Florida, and a member of the American Academy of Neurology, examined a group of Alzheimer's variants collectively known as atypical Alzheimer's disease. Rather than beginning with the memory problems most commonly associated with the condition, these variants strike other cognitive domains first: visual processing, language, executive planning, or coordinated movement.

The research specifically examined four such variants: posterior cortical atrophy, which primarily affects visual processing and spatial awareness; logopenic variant primary progressive aphasia, which disrupts language and word-finding ability; dysexecutive Alzheimer's disease, which impairs planning and organizational function; and corticobasal syndrome, which affects movement coordination alongside cognition. All four share the same underlying Alzheimer's pathology found in the more familiar memory-first form of the disease, but the order in which symptoms appear, and which brain regions bear the initial brunt of damage, differs considerably.

The gap the study actually measured

The researchers set out to determine whether people diagnosed with these atypical variants would actually qualify for newer anti-amyloid therapies, a class of disease-modifying treatments designed to slow Alzheimer's progression by clearing amyloid plaque buildup in the brain. These therapies are typically administered early in the disease course, often at the very first sign of memory-related cognitive decline, a detail that turns out to matter enormously for how the eligibility criteria are structured.

According to the study's findings, up to 85% of people with these atypical Alzheimer's variants did not meet current eligibility criteria for anti-amyloid treatment, even though many of them were still in the early stages of their disease overall. That figure represents a strikingly high exclusion rate for patients who, in terms of disease progression and underlying biology, closely resemble the memory-first patients these treatments were originally developed and tested to help.

Why the criteria themselves are the problem

The core issue, according to the study, traces back to how existing eligibility criteria for anti-amyloid therapies were originally constructed. Dr. Shir explained the underlying gap directly: "Atypical Alzheimer's disease is often underrecognized and underrepresented in clinical trials." Because the clinical trials that established these treatments' safety and efficacy profiles, and the eligibility rules that followed from those trials, were built primarily around patients presenting with classic memory-first symptoms, the specific diagnostic and severity benchmarks used to determine eligibility don't map cleanly onto how atypical variants actually progress.

A patient with posterior cortical atrophy, for example, might show severe, disease-defining visual processing deficits while still performing relatively well on standard memory assessments, the very tests eligibility criteria often lean on heavily to gauge disease stage and severity. Under current criteria built around memory-based benchmarks, that patient's genuine, active neurodegeneration in a different cognitive domain may not register as severe enough, or may not be measured using the right assessment tools altogether, to meet treatment eligibility thresholds, even though the underlying disease process is just as real and just as amyloid-driven as it would be in a memory-first patient at a comparable stage.

Why this exclusion carries real clinical weight

The stakes here are not abstract. Anti-amyloid therapies represent one of the few available disease-modifying treatment options for Alzheimer's disease currently on the market, as opposed to medications that only manage symptoms without addressing the underlying pathological process. For patients who do qualify, these treatments offer a chance to meaningfully slow disease progression during the window when intervention appears to matter most. A systematic eligibility gap that disproportionately excludes an entire category of Alzheimer's patients, purely because their symptoms present in language, vision, or executive function rather than memory, effectively denies that group access to treatment based on which brain region their disease happens to strike first rather than on how severe or treatable their underlying condition actually is.

That distinction becomes especially significant given that atypical Alzheimer's variants, while individually less common than the classic memory-first form, collectively represent a meaningful share of overall Alzheimer's cases, and tend to affect somewhat younger patients on average, a population that arguably has even more to gain from any treatment capable of extending their functional independence.

A call to rebuild the rules around a wider range of symptoms

The study's authors frame their findings not simply as documentation of a problem, but as a direct call for revising how eligibility criteria are constructed going forward. Their recommendation centers on ensuring that treatment criteria for anti-amyloid therapies better reflect the full range of ways Alzheimer's disease can actually present clinically, rather than continuing to rely primarily on benchmarks built around the classic memory-first pattern that, while common, does not capture the disease's full diagnostic reality.

That kind of criteria revision would likely require incorporating diagnostic and severity assessment tools specifically validated for atypical presentations, things like standardized visual processing assessments for posterior cortical atrophy or language-specific severity scales for logopenic aphasia, rather than continuing to lean primarily on memory-oriented cognitive testing that simply isn't designed to capture where these patients' disease burden actually concentrates.

Why broader clinical trial representation matters going forward

Beyond the immediate eligibility question, the study's findings point to a deeper, longer-standing problem in Alzheimer's research: atypical variants have historically been underrepresented in the clinical trials that generate the evidence base guiding treatment decisions in the first place. That underrepresentation creates something of a self-reinforcing cycle, since eligibility criteria built from trials that included few atypical patients will naturally tend to poorly capture what eligibility should look like for that same underrepresented population once the drugs reach real-world clinical practice.

Breaking that cycle would likely require future Alzheimer's drug trials to deliberately recruit and adequately represent patients with posterior cortical atrophy, logopenic aphasia, dysexecutive presentation, and corticobasal syndrome, rather than treating these variants as a secondary consideration addressed only after a treatment's core eligibility framework has already been established around memory-first patients. Until that broader representation becomes standard practice, this study's findings suggest a meaningful share of Alzheimer's patients will likely continue facing treatment eligibility criteria built around a version of the disease that isn't actually theirs.

---

*Sources cited in this article include the peer-reviewed study published August 5, 2026, in Neurology, the medical journal of the American Academy of Neurology, and reporting from Newswise and MedicalXpress featuring statements from study author Dr. Dror Shir of the Mayo Clinic. All figures reflect reporting available as of August 5, 2026.*