Dealing with Elderly Parents Who Refuse Help
"No" from an aging parent is rarely about the specific help being offered. It's about what accepting it seems to mean: that they're no longer capable, that someone else is now in charge, that the decline they've been managing privately is now visible to their own children. Refusing the help is often the only lever of control left in the conversation, so people use it.
Direct requests tend to fail for exactly that reason; they ask for compliance in a moment already dominated by a fight to preserve autonomy. What tends to work better is making the ask smaller and reversible: not "you need a home health aide" but "can we try someone coming twice a week for a month and see how it feels." A trial is easier to agree to than a permanent change, and most trials that go well quietly become permanent on their own.
A neutral third party changes the dynamic more than most family conversations can. A parent who won't hear "you should stop driving" from an adult child will sometimes hear the identical sentence from their own physician, because it doesn't carry the same charge of a child overturning a parent. Bringing concerns to a doctor ahead of an appointment, so the doctor raises it, is a legitimate strategy, not a workaround.
It's also worth being honest about what refusal sometimes is: not stubbornness but early cognitive decline that makes the risk itself hard to perceive, or depression that makes engaging with any change feel pointless. If refusal is sudden, uncharacteristic, or paired with confusion, that's a conversation for a doctor before it's a conversation about the specific help being refused.
Some refusals never resolve into agreement, and the honest version of this advice includes that outcome. What's still worth doing regardless is making sure the practical information (accounts, providers, documents) is gathered even without full cooperation, because a parent's willingness to accept help and a family's need to be ready if something happens suddenly are two different problems.