The United States faces a demographic collision with no engineering fix on the horizon: dementia incidence will roughly double over the next 35 years while every pillar of the care system — workforce, funding, specialist pipeline — is contracting or stagnant. Two new studies in JAMA quantify the scale. Globally, new dementia cases will jump from 11.4 million in 2023 to 24.7 million by 2050. In the US alone, annual new diagnoses will rise from 514,000 in 2020 to 1 million by 2060. The drivers are structural, not mysterious. Life expectancy gains push more Americans into the age brackets where dementia concentrates. The Hispanic population, which nearly doubled between 2000 and 2024, faces 1.5 times the dementia risk of white Americans. Even if per-capita risk were reduced through prevention, the sheer volume of people entering high-risk age cohorts overwhelms the math. The care workforce is moving in the opposite direction. Almost all nursing homes reported difficulty recruiting staff in 2024, and half had limited new admissions. Nursing assistant and home health aide demand is projected to grow 48% and 42% respectively by 2035, yet median wages sit below $19 and $17 an hour. Retention is abysmal because the work — bathing, feeding, managing cognitive decline — is physically and emotionally brutal for poverty-adjacent pay. The specialist pipeline is worse. The US had 10,270 geriatricians in 2000; by 2022, just 7,413. Geriatrics is among the lowest-paid medical specialties, and residency interest has not kept pace. The people trained to manage the most complex cases of an escalating crisis are literally fewer in number than they were a quarter-century ago. Into this gap steps the federal government — moving backward. The One Big Beautiful Bill Act of 2025 enacted nearly $1 trillion in Medicaid cuts over 10 years. Medicaid pays for more than half of all long-term care spending in the US. States are trying piecemeal fixes — South Carolina's geriatric loan forgiveness, incentive payments for Medicaid home care workers in almost half of states — but these are band-aids on a structural wound. There are bright spots at the clinical edge. The FDA approved Kisunla and Leqembi, which modestly slow cognitive decline in early Alzheimer's, and new blood tests can help diagnose or even predict the disease. But as USC researcher Julie Zissimopoulos notes, these advances will themselves be costly for the healthcare system to deliver at scale — creating new demand on a system already failing to meet old demand. The honest framing is this: absent a cure, dementia is a care problem more than a medical one. NYU's Jordan Weiss is right that cardiovascular health and physical activity represent real prevention levers. But the infrastructure question — who provides years of increasing hands-on supervision, who pays for it, and who shows up for $17 an hour — remains unanswered. The default trajectory is that families absorb the cost, financially and emotionally, because the system was never designed to carry it.
