Autism and Neurotypical Relationships: Bridging Different Communication Styles
By Mark Ryan, BACP Accredited Psychotherapist and NCPS Accredited Relationship Therapist
One partner says the other never picks up on how they're feeling. The other says they're constantly guessing at unstated rules nobody will spell out. Both are describing the same relationship, and both are right. When one partner is autistic and the other isn't, this is one of the most common shapes a couple's difficulty takes, and it's usually framed as one person's problem to fix. The research points somewhere more useful, and considerably fairer to both of you.
This piece is written for either side of that relationship, the autistic partner who's tired of being cast as the one who needs to change, and the non-autistic partner who's trying hard and still getting it wrong more often than either of you would like. Neither position is comfortable, and neither person is failing at something straightforward. What's happening is closer to two people speaking different first languages and mistaking that for one of them refusing to learn the other's.
A note on language before going further. This piece mostly uses ‘autistic’ and ‘autistic partner’ rather than ‘person with autism’, and sometimes ‘on the autism spectrum’. That is a deliberate choice. In the largest UK survey of how people describe autism, ‘autistic’ was endorsed by a large share of autistic adults, family members and parents, though by considerably fewer professionals, while ‘on the autism spectrum’ was among the terms accepted across every group (Kenny, Hattersley, Molins, Buckley, Povey and Pellicano, 2016). Many autistic people describe autism as part of how they think and experience the world rather than something attached to them, which is why ‘person with autism’ can feel like treating it as separable, or as something to apologise for. Others prefer person first wording, and ‘autistic’ can sound blunt to some readers. That reaction deserves respect rather than dismissal. In the therapy room, I use whichever words each person uses for themselves, and I would encourage any couple reading this to do the same.
The old story, and why it doesn't hold up
For decades, the working assumption in autism research was straightforward, autistic people struggle to read and respond to other people's emotional states, and that struggle explains most of the friction in a relationship with someone who isn't autistic. The person expected to adapt was almost always the autistic one.
This assumption shaped more than research papers. It shaped how autistic adults were taught to relate to others, often through social skills training aimed at making them appear more conventionally expressive or attentive, with comparatively little attention paid to whether non-autistic people might also benefit from learning to read an autistic communication style more accurately. The burden of translation sat almost entirely on one side of the relationship, which is a hard place to be asked to do all the adapting from, particularly inside a relationship meant to be a source of rest rather than another arena for constant performance.
For many adults on the autism spectrum, this has meant years of masking, consciously performing eye contact, small talk, and expected emotional reactions that don't come naturally, specifically inside romantic relationships where the stakes of getting it wrong feel highest. Masking can help a relationship function day to day, and it also has a real, documented cost in exhaustion and a sense of not being fully known by the person closest to you, precisely the person a relationship is meant to feel safest being fully known by. A framework that only ever asks one partner to keep performing this work indefinitely isn't a stable long-term foundation for either person.
An alternative framework, now widely used across autism research and clinical practice, challenges this directly. The double empathy problem, first proposed by the autistic researcher Damian Milton, argues that miscommunication between an autistic and a non-autistic person is a mismatch running in both directions, not a one-sided deficit located in the autistic partner (Milton, 2012). Two people who process and communicate social information differently will struggle to read each other accurately, and the non-autistic partner is just as capable of misreading the autistic one as the reverse.
This reframe matters beyond being a fairer way to describe what's happening. If the difficulty is framed as one person's deficit, the natural response is to try to fix that person, more eye contact, more small talk, more of whatever the non-autistic partner finds intuitive. If the difficulty is framed as a genuine two-way mismatch, the natural response changes entirely, both people learning to read the other's specific style more accurately, rather than one person being quietly asked to become someone else.
What recent research adds
It would be too tidy to say the research now proves the mismatch is perfectly symmetrical. A large, very recent study using a naturalistic empathic accuracy test found real support for the double empathy framework, and something more textured underneath it, autistic participants were, if anything, somewhat better at accurately reading the emotions of other autistic people than non-autistic participants were, while also tending to underestimate their own empathic ability (Rum, Feldman, Genzer, Allison, Golan, Perry and Baron-Cohen, 2026). The authors were explicit that individual variation was large enough to caution against treating either neurotype as a uniform category. Some of the mismatch in a mixed-neurotype relationship is structural, and some of it is simply two specific people, whatever their neurotype, learning to read each other better over time.
The finding that autistic participants underestimated their own empathic accuracy deserves attention in its own right. Years of being told, directly or indirectly, that empathy is something you struggle with can shape how you interpret your own performance long after the evidence has stopped supporting that story. A partner who's internalised this may discount real moments of accurate emotional attunement as flukes, or assume a partner's positive response is generosity rather than an accurate read. Naming that pattern explicitly, out loud, between the two of you, can matter as much as any communication technique either partner goes on to learn.
What this means for relationship satisfaction specifically
A study looking directly at long-term relationships involving an autistic partner found something useful for a couple trying to work out where to put their energy. Partner responsiveness, how well each person identified and met the other's specific needs, predicted relationship satisfaction for both the autistic and the non-autistic partner, more consistently than autistic traits themselves did (Yew, Hooley and Stokes, 2023). In practical terms, the more promising target isn't making one partner behave more like the other. It's both partners getting more accurate and more explicit about what the other one needs, since that's the variable the research says is doing the real work.
This finding matters precisely because it points at something learnable rather than something fixed. Neurotype itself isn't something either partner can change, and research pointing at it as the central variable would leave a mixed-neurotype couple with fairly little to work with. Responsiveness is different. It's a skill both partners can build regardless of neurotype, and it applies symmetrically, the autistic partner learning to state their needs more explicitly rather than assuming they're obvious, and the non-autistic partner learning to ask directly rather than expecting intuition to fill the gap, a two-sided project rather than one partner's homework.
Partner responsiveness, how well each person identified and met the other's specific needs, predicted relationship satisfaction for both the autistic and the non-autistic partner
Where the specific friction tends to show up
A few concrete patterns come up often enough to name directly. Directness that reads as blunt or even harsh to a non-autistic partner is frequently just clear, literal communication, meant exactly as stated, without an implied layer underneath it. A need for explicit information, what exactly is wrong, what exactly would help, can look like an autistic partner failing to intuit something a non-autistic partner assumes is obvious, when it's a reasonable, honest request not to have to guess.
Sensory needs and processing time cause a different kind of friction. A partner who needs quiet after a demanding day, or who needs a few extra seconds before responding in a conversation, isn't withdrawing from the relationship. They're managing a nervous system that works differently under load, and reading that management as emotional distance is one of the most common, and most fixable, misreadings in a mixed-neurotype relationship.
A third pattern shows up around unstated social expectations specifically. A non-autistic partner might expect a birthday, an anniversary, or a difficult day at work to be marked in a particular, unspoken way, flowers, a specific tone of concern, a particular kind of check-in, without ever stating the expectation directly. When that unstated script isn't followed, it can read as not caring, when the more accurate explanation is that the script itself was never shared. A partner on the autism spectrum asked directly, tell me exactly what would help right now, will often meet that request precisely and warmly. The same partner left to infer an unstated convention may simply never locate it.
The same pattern runs in reverse too, and it's the direction non-autistic partners tend to notice less in themselves. A direct request from a partner on the autism spectrum, an explicit description of a sensory limit, a plainly stated need, can be met with a non-autistic partner's own unstated assumption that a request phrased so plainly must be less emotionally significant than one wrapped in softer, more indirect language. The directness isn't a sign the need matters less. It's simply how the request was made, and reading tone or diplomacy as a proxy for importance is its own kind of misreading, one that can leave a clearly stated need quietly deprioritised precisely because it was easy to understand rather than hard.
What tends to help
Naming the double empathy framing directly, out loud, to each other, tends to do more good than either partner privately concluding the other one is careless or cold. It reframes both people's confusion as a shared, solvable translation problem rather than a verdict on either person's character.
This tends to change the content of an argument, not just its tone. Instead of he never notices how I feel or she's impossible to read, the conversation can become something closer to I think we're reading this situation differently, can you tell me what you noticed and I'll tell you what I noticed. That's a different conversation to have, one aimed at comparing two accounts rather than establishing which partner failed, and it tends to defuse the defensiveness that a blame-first framing almost always produces on both sides.
Being explicit rather than relying on implication helps both partners, not just the autistic one, though it often needs to be agreed on deliberately rather than assumed. Structuring a difficult conversation rather than hoping the right moment and the right words arrive naturally tends to work better here than in most relationships, precisely because it removes the guesswork the double empathy problem shows is difficult for both of you, not a failure specific to either side.
Where a specific difficulty has been misread as not caring or not trying, it's useful to check whether blame has quietly attached itself to something that's simply a communication style difference, since that reframing alone often does real work before any practical system is even in place.
Where distance has built up over time, that particular kind of loneliness inside a relationship deserves its own honest conversation, separate from whatever communication adjustments get made going forward.
How therapy can help
Couples navigating this often arrive already aware that something structural is at play, without a clear sense of what to do about it day to day. Useful work usually starts by mapping the specific pattern, which situations produce the most misreading, what each partner needs in those moments, and what's already being assumed rather than stated.
Where an autism diagnosis has arrived in adulthood, sometimes after years of a relationship already shaped by unexplained friction, there's often a real, specific grief mixed in alongside the relief of finally having a framework. Both partners may need space to look back at years of arguments through this new lens, recognising some of what happened as the mismatch this piece describes rather than either partner's failure, without it turning into a wholesale rewriting of the relationship's history.
This reinterpretation needs to hold two things at once to be useful rather than simply comforting. The new framework can explain a great deal of what felt inexplicable before, and it doesn't retroactively erase moments where one partner was, in fact, unkind, dismissive, or simply not trying. A diagnosis is a real, useful lens, not a blanket amnesty for every difficult moment in the relationship's history, and conflating the two tends to leave old, legitimate hurt unaddressed under the cover of a tidier explanation.
This work isn't about coaching the autistic partner to appear more neurotypical, or asking the non-autistic partner to simply try harder to empathise. It's closer to both partners learning a second, shared language for a relationship that already has two native ones, built around what the research points to, mutual translation, not one-sided adaptation.
In practice, this often means building specific, agreed tools together rather than relying on either partner's instincts under pressure, a shared vocabulary for naming sensory overload before it becomes withdrawal, an agreed way of asking directly for what would help rather than waiting to be understood, a habit of checking a read before acting on it rather than assuming it's correct. None of these are complicated on their own. What makes them useful is building them together, deliberately and with real care, rather than each partner privately developing workarounds the other one never gets to see or benefit from.
If you feel like now might be the right time to talk this through with a therapist, I offer a free 30-minute consultation in person (in London) or online. Either partner is welcome to reach out first, and you don't need to already agree on what the underlying pattern is before starting, that's often exactly the kind of thing this work helps clarify together.
About the Author
Mark Ryan is a BACP Accredited Psychotherapist and NCPS Accredited Relationship Therapist working in person across central London, from rooms in Pimlico, Kensington, and Angel. He works with couples, individuals, and polycules on relationship issues including jealousy, affairs and betrayals, sex problems, and considered separation. His practice is integrative, sex-positive, and affirming of all relationship structures and identities.
References
Milton, D. E. M. (2012). On the ontological status of autism: The ‘double empathy problem’. Disability & Society, 27(6), 883–887. https://doi.org/10.1080/09687599.2012.710008
Rum, Y., Feldman, N., Genzer, S., Allison, C., Golan, O., Perry, A., & Baron-Cohen, S. (2026). Nuances of double empathy in autistic and non-autistic people: Examination using the empathic accuracy paradigm. Molecular Autism, 17, 34. https://doi.org/10.1186/s13229-026-00723-2
Yew, R. Y., Hooley, M., & Stokes, M. A. (2023). Factors of relationship satisfaction for autistic and non-autistic partners in long-term relationships. Autism, 27(8), 2438–2450. https://doi.org/10.1177/13623613231160244
Kenny, L., Hattersley, C., Molins, B., Buckley, C., Povey, C., & Pellicano, E. (2016). Which terms should be used to describe autism? Perspectives from the UK autism community. Autism, 20(4), 442–462. https://doi.org/10.1177/1362361315588200
