How effective are cholinesterase inhibitors in Alzheimer's disease?
These drugs show a measurable slowing of mental decline in Alzheimer's disease, although the effect is modest. Galantamine and donepezil have the strongest evidence base; discuss the options with a neurologist or treating physician.
Donepezil, galantamine and rivastigmine slow cognitive decline in Alzheimer's disease, but the effect is modest. On cognition tests, people taking these medicines score on average 2 to 3 points better than people taking a dummy pill. That just clears the threshold at which researchers consider a difference clinically meaningful, but it does not resolve memory problems.
Galantamine has the broadest support. At a daily dose of 16 to 24 mg, it improves not only cognition after six months but also the ability to carry out everyday activities and behavioural symptoms such as agitation. Notably, participants who received galantamine had a lower risk of death after six months than those in the dummy-pill group (1.3% versus 2.3%), although the absolute numbers were small. In people with mild cognitive impairment (an early stage before Alzheimer's disease) galantamine does not work: after two years there was no difference whatsoever compared with the dummy-pill group.
Donepezil (10 mg/day) shows small but measurable improvements in cognition and everyday activities in mild to severe Alzheimer's disease. On behavioural symptoms and quality of life, donepezil makes no demonstrable difference. A higher dose of 23 mg offers no additional benefit but does cause more side effects and more drop-outs. There is also a signal from smaller studies that donepezil slightly slows the shrinkage of memory regions in the brain, but that finding is preliminary.
Side effects are a genuine concern. With galantamine, more than one in five users stop treatment early, and nausea occurs in 21% compared with 8% in the dummy-pill group. Donepezil causes side effects in 72% of users versus 65% in the dummy-pill group. Newer medicines based on targeted antibodies (such as lecanemab) slow decline slightly more than these drugs, but the difference does not reach the threshold that would matter in everyday life, while the risk of serious brain abnormalities in people who carry a specific inherited risk gene (APOE4) is more than five times higher.
This answer is based on seven pooled analyses of multiple studies in people. The researchers themselves rate part of the evidence as high certainty (thinking ability, side effects, and death rates with galantamine) but also as low certainty (overall day-to-day functioning with galantamine). Four studies were flagged by their authors as carrying a high risk of distortion due to drop-outs or selective reporting. The evidence is strongest for galantamine and donepezil; for rivastigmine the supporting evidence is thinner. For people with early memory problems that have not yet reached Alzheimer's disease, there is no evidence that these drugs help.