Blood tests could transform Alzheimer's diagnosis, but real-world barriers remain

Alzheimer's disease (AD) accounts for roughly 60%–80% of dementia cases and is the seventh leading cause of death worldwide. Yet diagnosis still leans on symptoms that appear once cognitive impairment already interferes with daily life. The authors note that 25%–35% of patients clinically diagnosed with AD in specialized clinics are misdiagnosed, and the rate is likely higher in primary care, where most people are first evaluated. Timely, accurate diagnosis matters because it allows better treatment strategy, lower societal and caregiver costs, and delayed institutionalization.

Cerebrospinal fluid (CSF) and PET biomarkers can detect AD pathology, but they are rarely used in routine care. Lumbar puncture is seen as invasive, PET involves ionizing radiation and high cost, and both need appropriately equipped facilities. High-performing blood-based biomarker (BBM) tests reach sensitivity and specificity of about 90%, making them a potentially cost-effective, minimally invasive alternative. This Perspective is not a new experiment. It reports an international roundtable of 10 experienced AD clinicians and biomarker experts from China, Germany, Japan, the Netherlands, Spain, Sweden and the United States.

The roundtable proposed that BBMs enter practice in two phases. In phase one, specialists would use two-cut-point BBM tests that give a high positive predictive value and a high negative predictive value, with an intermediate range between the cut-points. Initially, BBM-based diagnoses would be confirmed by CSF or PET, regardless of whether the result falls in the intermediate range. Shared decision-making between patient and clinician could add further validation. Results must always be interpreted within the full clinical context. In phase two, the same pathway would run routinely in non-specialist settings to streamline referrals to specialist centers.

The authors cite a US health-system simulation by Mattke and colleagues. Referring patients using the Mini-Mental State Examination (MMSE) alone or BBMs alone would increase specialist appointments and waitlists. Combining BBMs with the MMSE would shorten waiting times and potentially reduce costs, letting specialists focus on people with confirmed or highly likely AD.

The experts identified three main barriers: limited real-world data, resource gaps and system barriers. Data are thin for cognitively unimpaired people, for a wider range of dementia diagnoses, for patients with chronic kidney disease, obesity or other chronic conditions, and for racially and ethnically diverse populations. This matters because the prevalence of amyloid pathology in a population affects test performance. Prospective real-world studies testing non-inferiority against CSF and PET are also scarce.

Resource gaps are real. Non-specialist providers are often overburdened with little time per patient. Some consider CSF or PET testing unreasonable because of invasiveness, cost and perceived low benefit, and some family physicians may skip the initial cognitive assessment altogether. The paper is expert opinion, so it offers no effect sizes or p-values. Several authors are employed by Eli Lilly, which a reader should weigh. The authors conclude that BBMs could streamline diagnosis, but more real-world evidence and multi-stakeholder collaboration are needed first.