The second person in the room: late diagnosis and relationship outcomes

Adults are now being diagnosed autistic or ADHD in their thirties, forties and fifties. The average age at an adult autism diagnosis sits somewhere between 31 and 49. ¹ By that age, most people have a partner, and have had one for a long time. So a late diagnosis doesn’t land in a life on its own. It lands in a relationship that was built, argued over and held together for years without it.

Does the relationship change after that? Does it get better or worse, for how long, and for which partner first?

1 What my book couldn’t prove

When I wrote Loving Variant Minds, I gave a whole chapter to relationships, and to what happens to a couple after one of them is diagnosed. It tends to go something like this. One partner gets a new way of understanding themselves, usually quickly and often obsessively. The other partner gets it in bits and pieces. Both of them have to go back over years of shared history. Roles that were settled come up for grabs again. Old arguments get reopened, this time with new evidence. And the explanation that was supposed to make everything easier makes things harder, at least for a while.

That’s what clinicians and couples describe. What I couldn’t do was point to a single study that had followed couples through it and measured what actually happened. I looked then, I’ve looked again since, and I still haven’t found one. ² So that chapter rests on clinical observation and people’s own accounts. That bothered me then, and it bothers me more now.

2 Divorce rates can’t answer it

The obvious place to look is divorce rates. It’s also the wrong place, and the mistake is everywhere in popular writing on this.

Divorce only counts people who got married in the first place, and autistic and ADHD adults don’t marry at the same rate. In a study of 306 adults grouped by their autism and ADHD traits, the group with more autism traits had fewer married people and more who had never married, while the higher divorce rate turned up in the ADHD-traits group. ³ So the comparison groups aren’t the same to begin with. If autistic adults marry later and more selectively, the ones who do marry may already be the strongest couples. A low divorce rate among them could mean good matches, or it could just mean the shakier relationships never became marriages and were never counted.

Divorce also turns something gradual into a yes-or-no event. It can’t see the marriage that carries on while both people are miserable - and that’s exactly the marriage I’m asking about. Fewer people marry now, and de facto relationships aren’t registered anywhere. Relationships end in other ways too: one partner dies, or the two drift apart and keep sharing an address. And a divorce can be finalised years after the relationship actually ended, which makes it almost useless for working out whether something changed after a diagnosis.

3 What we know about neurodivergent couples

There is research on couples where one or both partners are autistic. It’s small, but it’s real.

Yew, Hooley and Stokes surveyed 95 autistic adults and 65 non-autistic partners. For both, feeling understood and cared for by their partner predicted how satisfied they were with the relationship. ⁴ That takes the focus off the autistic partner as the only thing that matters. Khaw and Vernon compared 106 autistic adults by their partner’s neurotype and found no significant difference in satisfaction between autistic-autistic, autistic-neurotypical and other neurodivergent pairings, although what helped and what got in the way differed. ⁵ Smith and colleagues talked to couples about what keeps these relationships going. ⁶ And Attwood and Aston, from their clinical work and a survey, suggest autistic-autistic couples may be more satisfied, although few of those couples responded. ⁷

None of these studies asks when the diagnosis came. People are recruited as diagnosed autistic adults and their partners. Whether the diagnosis arrived at seven or at 47, before the relationship started or fifteen years into it, isn’t asked, reported or analysed.

4 What we know about late diagnosis

The research on late diagnosis is bigger, and growing fast. A 2026 review of 25 studies on how adults rebuild their sense of self after an autism or ADHD diagnosis found relief, mixed with grief, frustration and disorientation, and concluded that support after diagnosis has to deal with identity. ⁸ Older adults diagnosed late in life describe their world being turned upside down, and the researchers argue a diagnosis should be the start of the conversation, not the end of it. ⁹ A study of 19 autistic adults and four of their support people found that informal support - which usually means a partner - helped, while also putting strain on the very relationships providing it. ¹⁰

Here, the relationship turns up as background, as a source of support, and sometimes as collateral damage. It’s never what’s being measured. Nobody asks both partners, and nobody follows the couple over time.

The closest thing I’ve found to my question is a small qualitative study by Holmes on later-in-life diagnosis and marital satisfaction. ¹¹ It’s one study.

So there’s one body of research that looks at couples and ignores timing, and another that looks at timing and ignores couples. My question falls in the gap between them.

5 Which came first?

This is the part that makes the question hard to answer.

Adults don’t seek an assessment out of the blue. Something pushes them, and trouble at home is one of the most common pushes. A study of 160 adults referred for assessment found that people on the ADHD pathway were usually responding to a crisis and wanted their difficulties recognised, while people on the autism pathway wanted a way to understand themselves - and to make sense of what was going wrong in their relationships and at work. ¹²

So if relationships look worse after a late diagnosis, that might have nothing to do with the diagnosis. It might just be the same trouble that sent the person for assessment in the first place. Asking people afterwards doesn’t fix that, because everyone you ask has already been through it.

There’s an interesting group here that nobody seems to talk about: the people who are assessed and not diagnosed. In one UK service, that was about one in ten of those referred for autism. ¹ They went through the same doubts and the same strain, often the same long wait - a median of 252 days in one Scottish audit ¹³ - and then got a different answer. If anyone could show whether the diagnosis itself changes a relationship, it would be by comparing those two groups.

And it’s possible the diagnosis makes no difference at all, once you allow for whatever prompted the assessment. The upheaval couples describe could be real, and still fade out over a few years, or turn out to be no different from the ordinary strain relationships go through in midlife. That would be worth knowing too.

What I think

My guess is that the first year after a late diagnosis is harder than the year before it, and that things mostly settle after that. I also think the two partners go through it at different speeds - the diagnosed partner racing ahead, the other one catching up - and that the gap between them matters more than how either one is doing on average. But that’s a guess, built from what people tell me and what I wrote in the book.

My own marriage doesn’t fit that guess, because it didn’t make it to the diagnosis.

It ended earlier, at the point where I started to recognise my own masking for what it was. Once I saw it, I had an almost physical need to stop - and stopping took me into a shutdown. Autistic adults describe this kind of burnout as deep exhaustion and a loss of skills they used to have, and they name unmasking as part of how they recover from it. ¹⁴ That’s what it was for me. To my spouse, it looked like I’d given up - a loss of hope, at exactly the time we most needed each other.

The road there didn’t help. My path to a diagnosis went through other diagnoses first, and each one that turned out not quite right made the next one harder to believe, for both of us. By the time autism and ADHD came up, they sounded like one more false start.

So for me, the hardest stretch came before the diagnosis. I doubt I’m the only one. If the damage starts during the discovery - the unmasking, the wrong turns, the shutdown - then anyone who only looks at what happens after the diagnosis will miss it. It’s the chicken-and-egg problem again - the strain often comes first.

What I’m sure of is that there’s almost no support after an adult diagnosis, and what little there is is aimed at the person diagnosed. A clinician hands over a diagnosis that reframes forty years of someone’s life, and then the appointment ends. The person goes home to a partner who wasn’t in the room, hasn’t read anything, and is about to spend the next year hearing that things they took as unkindness were neurology. There’s no leaflet for that, and no session for it either. Neither of them has anything to tell them what’s normal, how long it lasts, or whether they’re watching a relationship fall apart or just adjust.

That gap is a big part of why I wrote Loving Variant Minds. I’m coming into psychology from outside it, and I see its main tool as explanation more than cure - helping people understand what’s happening to them, and giving them a way through it. A book was the only way I could see to put something between a newly diagnosed person and the partner trying to understand them. But a book built from clinical accounts and people’s stories can only go so far. I could describe the pattern. I couldn’t tell anyone how common it is, how long it lasts, or which couples come through it, because nobody can.

If the pattern turns out to be predictable, it can be prepared for. That doesn’t need to cost much. It could be one conversation at the point of diagnosis, with the partner there: the next twelve months will probably be harder than the last twelve, this is why, this is what tends to help, and this is what it means if things haven’t settled by then. From what happened to me, I’d want that conversation to start earlier too - when someone is referred for assessment, not just when the answer comes back. Nobody can have that conversation yet, because nobody knows what to say.

⁂

---