The earliest signs of Alzheimer disease frequently do not appear in a doctor office. They appear in a bank statement. Researchers have now documented a consistent pattern across multiple datasets: people with early-stage dementia and mild cognitive impairment begin making detectable changes in their financial behavior years — sometimes a decade — before a clinical diagnosis is made. Missed payments on accounts that have never been missed. Investment decisions that deviate from established patterns. Elevated vulnerability to financial scams. The financial record is longitudinal by nature. The signal is there. Medicine has been slow to read it.
The mechanism is not obscure. The prefrontal cortex governs executive function, financial planning, and impulse control. It is among the first regions compromised in Alzheimer disease, vascular dementia, and Lewy body dementia. Behavioral changes in that domain can precede the memory impairment that triggers clinical evaluation by years. Cognitive screening tests — the Mini-Mental State Examination, the Montreal Cognitive Assessment — are designed to catch impairment that is already substantial. They are, by design, late-stage instruments. The financial record can function as an earlier one.
The case for building detection systems around this signal is straightforward on the clinical side. Earlier diagnosis expands the window for care planning, legal preparation, and treatment decisions made while the patient retains full capacity. Dementia costs the United States an estimated three hundred sixty billion dollars annually in direct care. Earlier diagnosis compresses that cost curve by shifting care toward less expensive earlier-stage management. The data exists in financial institutions already. The pattern has been replicated. The detection gap is institutional and regulatory, not evidentiary.
The case for caution is equally serious and should not be dismissed in the enthusiasm for a promising signal. Financial behavior changes for many reasons unrelated to cognitive decline — job loss, caregiving obligations, health costs, relationship change, market conditions. A system that treats missed payments as a dementia signal will produce false positives that carry stigma, affect credit, and potentially trigger legal proceedings around financial capacity that the affected person cannot easily contest. Building any clinical or financial surveillance architecture around this pattern requires consent frameworks, privacy protections, and appeals processes that do not currently exist and would need to be built deliberately before deployment.
The distributional question is who captures the benefit of earlier detection and who bears the risk of misidentification. Earlier diagnosis benefits patients and families most directly — more time to plan, more decisions made with capacity intact. The risk of false positive identification, and its financial and legal consequences, also falls to patients and families, disproportionately to those without resources to contest an institutional determination that labels them impaired. The healthcare system captures system-level cost savings that do not automatically translate into individual benefit without policy design that routes those gains back to the people the system is supposed to serve.
The standard of care for dementia detection has not changed substantially in decades. It relies on patient or family symptom report, clinical observation, and screening tools designed for advanced impairment. The financial behavior signal represents an earlier layer in what could become a more complete detection system. Whether it becomes that depends on whether medicine, financial institutions, and regulators can build the consent and protection architecture required before deploying the capability they already technically have. That architecture does not yet exist. Building it is the work.
American Life is independent, advertisement-free analysis.
