Stopping a Statin in Advanced Dementia: What a Decade-Old Trial Actually Tested
A statin taken faithfully for years after a distant heart attack, now sitting on the medication list of a woman with advanced dementia and a prognosis measured in months. A real trial tested exactly this decision — the question is whether her own history places her inside it.
E.S., an 84-year-old woman, taught third grade for thirty-two years before retiring, and for a long time after that still volunteered reading to students at her old school on Friday mornings — a routine that stopped four years ago as her memory began to fail in ways that were no longer explainable by normal aging. She now has advanced Alzheimer's dementia: she no longer recognizes her own daughter reliably, requires full assistance with eating and dressing, and her care team's honest estimate of her remaining prognosis is measured in months. She had a myocardial infarction a decade ago, well before her cognitive decline began, and has taken atorvastatin 40mg daily for secondary prevention ever since without complication. Her daughter, who holds healthcare power of attorney, brought up stopping the statin at today's visit herself — not because of any new problem, but because she has started to wonder, watching the daily struggle to get her mother to swallow a full handful of pills, whether this particular one is still doing anything for her.
Kutner and colleagues answered a version of this exact question directly in a 2015 randomized trial: 381 patients on long-term statin therapy with an estimated life expectancy between one month and one year, drawn from palliative care settings, were randomized to discontinuation or continuation. Patients on a statin for a recent, active cardiovascular event were excluded, but both primary- and secondary-prevention patients — those, like E.S., whose statin was protecting against a real prior event rather than an unrealized risk — were included. There was no significant difference in survival between the two groups, and the discontinuation group scored better on quality-of-life measures and had lower medication costs. Statins also have a genuinely long time-to-benefit in the trials that established their value in the first place — mortality curves typically take one to two years or more to separate — a timeline that sits uneasily against E.S.'s own realistic prognosis regardless of what the discontinuation trial found.
Geriatrics clinic, medication review
Kutner's trial tested this exact question and found no survival difference, with better quality of life and lower burden in the discontinuation group. Statins also take years to show a mortality benefit in the trials that established them — years E.S.'s own prognosis doesn't realistically include. I'd stop it.
This isn't a reason to stop everything reflexively — it's specific to a drug whose whole value proposition is years away.
I'd be cautious about extending that reassurance to her specifically. She's on this for secondary prevention after a real MI, not primary prevention against a risk that never materialized — and secondary-prevention trials have historically shown earlier separation than primary-prevention ones. I'm not sure the safety data translates as cleanly to a patient whose statin is protecting against a proven prior event.
Actually — checking Kutner's own enrollment criteria as I say this, the trial included secondary-prevention patients directly, excluding only those with a recent, active event. Her MI is a decade old. That's the population the trial actually studied, not a population I can argue she falls outside of.
Agreed — and worth saying plainly to the family: stopping this medication isn't giving up on her care, it's recognizing that this specific pill's entire reason for existing was a multi-year bet on cardiovascular risk that her prognosis doesn't leave room to collect on. The daughter's own instinct, watching the daily struggle to administer it, was pointing at something real.
Agreed: discontinue atorvastatin today, frame the decision to the family explicitly as burden reduction rather than a withdrawal of care, and schedule a broader medication review at the next visit rather than reconsidering the rest of her regimen today.
Not agreed, and carried forward explicitly rather than smoothed over:
Accepted that Kutner's trial population includes her, but noted this as a genuine update to his own initial read, not something he'd have assumed without checking — worth remembering as a habit for future secondary-prevention deprescribing questions.
All three voices agreed the decision would be revisited directly rather than assumed permanent, consistent with treating today's prognosis estimate as real but not fixed.
The daughter's own observation — that the daily struggle to administer the pill was itself worth naming — was treated as clinically relevant information, not just an emotional aside to be managed around.