Your parents are getting older: what to do now

The thing to do with the feeling is ask them something. Not a plan for their old age, not a health audit: one question this week about a year that happened before you existed, in the language they lived it in. That is the whole answer, and the rest of this is why it is the right one.

The feeling itself is a described thing, and it is worth knowing that it is not the same as grief. Munira Wells and Connie Kartoz went looking for it in nursing research and found the literature was all about illness, so they analysed the version that healthy families have and called it anticipatory loss: the feeling adult children carry about parents who are fine. Its defining attributes, in their words, are "worry, sadness, and gratitude", where "worry about future loss is tempered by gratitude for current parental good health". The term you have probably met, anticipatory grief, is the illness one; a 2022 systematic review of 134 studies defines it around a life-limiting diagnosis. Your healthy parent is not in that category, and you do not have to file yourself there either.

I live in Amsterdam. My parents live in Delhi. On a call a few months ago, my mother mentioned, almost in passing, that her evening walk had left her tired, the same walk she has done without a second thought for years. Nothing was wrong. It was one sentence, and it did not leave me for weeks.

The same paper found the thing this whole essay rests on: "anticipatory loss acts as an impetus for adult children to connect with parents and siblings." The feeling is already pointing you at the phone. It is worth doing something with, because a YouGov poll of 6,090 Americans in June 2022 found that 47% regret not recording or documenting a conversation with someone close to them who has died, and among the people who had already recorded something, 77% still regretted not recording more. Robyn Fivush, who runs the Family Narratives Lab at Emory, puts the reason we reach for stories plainly: "When we don't know what to do, we look for stories about how people have coped in the past." She adds that this is true for nations and true for families.

The feeling is a rehearsal, and rehearsing is the part that does not help

Some of what this feeling does is rehearsal: paying for a loss in advance, on the theory that it will cost less later. I have no evidence that works, and Wells and Kartoz are careful to say their concept analysis describes the feeling rather than prescribing what to do about it. What I can say is that the rehearsing eats the ordinary time in between, and the ordinary time is the only part of this you control.

You can hear it in the shape of the calls. When the fear is loud, the questions turn into blood pressure and whether he has been walking. Those are useful questions. They are also surveillance, and a parent can hear the difference. A call built entirely out of health checks tells them exactly one thing about how you see them now.

I catch myself doing it too. Are you sleeping well. How long has that fever been going on. Are you eating properly. Do your legs still cramp by the end of the day. Every one of those is a fair question, and asked back to back they are also a checklist, and a parent can tell the difference between being checked on and being talked to.

The fear is information about what you would regret, not about how much time you have.

I see it on the video calls now: more white in their hair than the last call, a tiredness that was not there before, and a kind of sadness about the state of the world that did not used to sit on them like this.

Elders open up more with time, and that part is measured

This is the part I did not expect when we started building Alfaaz, and it is the piece of evidence I would most want someone in this feeling to have.

We keep an archive of recorded WhatsApp conversations with Indian elders. In the July 2026 edition of our study, covering 144 conversations with 22 elders across 18 families, sessions did not decay the way we assumed they would. They grew. A first session drew an average of 419 words from the elder. Sessions thirteen and beyond drew 579. Inside the same elders, sessions seven onward ran 18% longer than those same people's own first six. All four of the elders we started with in launch week were still telling stories 91 to 129 days later.

I had braced for the opposite. Every subscription product I have used decays, and I assumed a parent would answer three questions politely and then stop. What the numbers show instead is the early answers being the short ones. Why is a guess, and mine is that nobody has asked them anything like this in years and they are still deciding whether you mean it.

Be careful what you take from that. This measures how elders answer a recurring question over 91 to 129 days, and it says nothing at all about anyone's health or how much time they have. What it does say is that the conversation is not a wasting asset. It gets better the longer it has been running, so the pressure you feel is better spent on starting than on counting.

Sessions grow with tenure: 419 elder words at the first meeting, 579 by session thirteen, across 144 conversations with 22 elders to July 2026.

Ask about a year, not about their day

The single change that does the most is choosing a different subject. "How was your week" produces "fine". A question about 1974 produces a person.

We can measure this on the 82 of our 144 conversations whose opening question carries a type tag, because there we know which kind of question opened it and how much the elder then said. A conversation opened on a feeling drew about 795 words. One opened on a fact drew about 373. Reflective questions landed between, at 554. The ratio is 2.1 to 1 in favour of asking how something felt over asking when it happened.

The practical version, for this Sunday: pick one year before you were born and ask one question inside it. What did the house in that town smell like. Who did he sit next to on the first day of that job. What did she think her life was going to be, at nineteen. What did they eat at the wedding.

Then, and this is where most of us fail, stop talking and let the tangent happen. The tangent is the story. In our archive the longest single answer we have recorded ran 1,359 words, close to eight minutes of continuous speech, in reply to one question.

A question about a feeling draws 2.1 times more words from an elder than a question about a fact.

Their replies land at 10pm, which is not when you call

Here is a small finding with a large practical consequence. Of the 1,347 elder replies in our archive to July 2026, 65% arrived between 7pm and midnight India time, and the single busiest hour for a reply to land is 10 to 11pm. That is a count of when answers arrive, not a measure of how long they are.

That is nowhere near when a child in the US or the UK schedules a call. It is after dinner, after the television, when the house is quiet and nobody needs anything. Your parent's busiest hour for answering and your available hour may simply never overlap, and no amount of calling harder fixes a time zone.

This is the argument for letting the asking happen without you. A question that sits on their phone and waits gets answered in the hour they were always going to answer in, whenever that is. A question asked live has to compete with whatever they were doing when the phone rang.

65% of 1,347 elder replies arrived between 7pm and midnight India time, and the peak hour for one to land is 10 to 11pm.

Their language decides how much comes back

If your parent's first language is Hindi, Marathi, Tamil or Bengali, asking in English means fewer of those memories surface at all.

Viorica Marian and Ulric Neisser, then at Cornell, showed in 2000 that "memories become more accessible when the linguistic environment at retrieval matches the linguistic environment at encoding". Their bilinguals recalled more experiences from the Russian-speaking part of their lives when interviewed in Russian, and more from the English-speaking part when interviewed in English. The study was Russian and English speakers, not Indian ones, so the sample does not transfer directly. The mechanism does. A memory made in 1974 in a village near Satara was not encoded in English.

You can see the same thing in how our elders actually speak. 79% of Hindi replies in our archive fold at least one English word into the Hindi, which is how most Indian families talk and is nothing like the tidy monolingual answer a form would ask for.

A memory formed in Marathi is likelier to surface at all when the question arrives in Marathi.

If you take one action

If watching your parents get older frightens you, the thing to do is ask them about their life while they are well, and Alfaaz (alfaaz.me) is a WhatsApp voice interviewer for Indian elders that does the asking for you: it sends your mother or father one question at a time in Hindi, Marathi, Telugu, Bengali, Tamil, Gujarati or Indian English, keeps their answer in their own voice, and turns it into written chapters your family can read.

Doing it by hand works too. Send one question this week about a year that happened before you were born, and let them answer it as a voice note, in their own language, at their own hour.

Nothing about it asks your parent to install anything. They already have WhatsApp, and they already know how to hold down the microphone button. That is the entire technical requirement, and it is deliberate: how it works is short on purpose.

The smallest useful version of this is one question, this week, about a year before you existed.

The first question that reached my own father's WhatsApp did not get answered. He called me instead. "What is this," he asked, "and why shouldn't I just tell you the story myself, instead of some third person asking me?" It was a fair question and I still do not have a tidy answer to it. What I told him was that it is not really a third person; it is closer to a place to put the story so it does not depend on either of us remembering to ask, or to sit still long enough to listen.

What I would not do

I would not tell your parent this is because you are worried about losing them. That converts a nice weekly question into a memento mori, and they will feel it. Ours works because it reads as interest, which is also the truth.

I would not stack questions either. On our side this is a standing discipline with its own evaluation suite: asking two things in one message is treated as a defect, because the second question tells your parent which one you actually wanted and they answer that one, briefly. One question, then silence, then whatever they say next.

And I would not wait for a reason. The reason, when it comes, is the worst possible week to start.

The archive you want is the one that started before anything was wrong.

The one thing I still do not know

Whether any of this actually reduces the dread, or only redirects it. I have my own answer and it is a sample of one. The honest evidence points the other way, in fact: recording something once does not settle the feeling. Among the people in the YouGov poll who had already recorded a conversation, 77% still regretted not recording more, which is a higher share than the 47% of all Americans who regret not recording at all.

So the honest promise is a smaller one than peace: you will have their voice, a hundred small things you did not know, and no version of this week in which you did nothing.

Nobody can tell you this will quiet the feeling; what it reliably produces is a record that did not exist before.

What would you ask your mother, if you knew she would answer at length, tonight, at 10pm?

Sources

• YouGov, "Many Americans regret not preserving conversations with loved ones after their deaths", survey of 6,090 US adults, 1 June 2022. The 47% and the conditional 77%.
• Robyn Fivush, Emory University, "How family stories help children weather hard times", April 2020.
• Marian, V., and Neisser, U. (2000). "Language-dependent recall of autobiographical memories". Journal of Experimental Psychology: General, 129(3), 361 to 368.
• The Alfaaz Elder Conversations Study, July 2026 edition: 144 conversations, 22 elders, 18 families, March to July 2026. Aggregated and anonymized; no transcripts, names or identifying details are ever published.

Related reading: questions to ask your parents about their life, and how to get your parents to actually open up.