Thesis. The mainstream medical reflex to “cure at all costs” in people over 70 frequently mistakes longevity for mercy. The most humane act a physician can sometimes offer is to not prolong a life—to pivot from rescue to relief, from extension to meaning (Cassell, 1982; Institute of Medicine, 2014).
The hard arithmetic of late life
Modern demography gives us a sobering baseline: in a large U.S. cohort, the lifetime risk of dementia from age 55 to 95 is ~42%, higher in women and Black Americans (Fang et al., 2025). Epidemiologists call this a competing-risk world: many of us who avoid death from heart disease, cancer, or infection will live long enough to encounter cognitive failure (Seshadri et al., 1997). If we dread dying with dementia, the only mathematical alternative is to die of something else first—and to do so without torment.
Evidence of “too much medicine”
The problem is not hypothetical. A systematic review of end-of-life hospital care found widespread use of non-beneficial treatments—from last-days chemotherapy to futile ICU admissions—affecting roughly a third of dying patients across studies (Cardona-Morrell et al., 2016). Clinicians themselves report cultural and systemic pressures that drive high-intensity care even when unlikely to help (Dzeng et al., 2023). Campaigns like Choosing Wisely emerged precisely to curb low-value tests and treatments in older adults (Rowe et al., 2020; American Geriatrics Society, n.d.).
“This review has confirmed widespread use of non-beneficial treatments at the end of life in acute hospitals.” — Cardona-Morrell et al. (2016).
Not giving everything is not giving up
Pivoting from automatic cure to proportionate care is not abandonment; it is good medicine. The Institute of Medicine concluded that a “patient-centered approach to care near the end of life should be a high national priority” (2014). When conversations clarify what matters—staying home, avoiding ICU delirium, keeping pain controlled—treatments change accordingly.
What this means after 70
After 70, aggressively “curing” every detectable abnormality can replace living with medicalization. Treatment thresholds that made sense at 50 may invert their risk–benefit profile at 75. A physician should ask: Will an intervention likely restore function the patient values? Or will it extend a trajectory that the patient finds worse than the alternative? Sometimes, the greatest favor a clinician can do is to not prolong a life that has shed most of what the patient calls living.
Bottom line. We all die of something. If we do not want to die of dementia, we must die of something else before it.