The word arrived before the science
Alvin van Asselt, Desiree Reekers and Yvette Roke published a scoping review in Neurodiversity in 2026 mapping what the scientific literature actually knows about Rejection Sensitive Dysphoria in autistic adults. The answer, in short, is: strikingly little, and what exists barely agrees with itself. Two of the three authors are autistic, which matters for how the review reads — it approaches the subject from inside the experience rather than peering in from clinical distance, and it takes the community’s account seriously as a starting point rather than a curiosity to be corrected.
The central fact the review documents is an inversion of the usual order. Normally a clinical construct is defined in the research literature, validated, and then filters out into public use. RSD ran the other way. The term was popularised by the psychiatrist William Dodson on the basis of clinical observation, and then taken up, expanded, and given detailed phenomenological texture by neurodivergent people on social media — autistic people very much included, many of whom do not identify with ADHD but recognise the experience precisely. The community has a word for it, a rich shared description of it, and a high degree of consensus about what it feels like. The research apparatus has almost none of that.
The numbers tell the gap plainly. The authors screened 1,285 records and found 12 eligible studies. Not one of them explicitly used the term RSD. None investigated responses to criticism, despite criticism being half of what the community description covers. None evaluated any intervention. RSD does not appear in the DSM or the ICD, and the direct scientific grounding for the original concept rests on clinical observation and a handful of case studies. So the situation the review captures is this: a precisely-named, widely-recognised, vividly-described experience, sitting almost entirely outside the formal evidence base meant to understand it.
This is worth pausing on, because it inverts the assumption that usually governs whose knowledge counts. The default hierarchy treats the clinical literature as the authoritative account and lay description as the unreliable shadow of it — something to be checked against the science before it can be trusted. Here the lay description came first, holds together better than the science does, and is the thing the science is now scrambling to catch up to. Over half the included studies were published in 2024 or 2025, a research field hurrying to investigate something the people living it have been describing in detail for years. The autistic community did the conceptual work. The apparatus is the one playing catch-up.
That ordering is the first thing this review establishes, and it is the part most easily lost when RSD gets discussed as though the open scientific questions mean the experience itself is in doubt. The experience is not in doubt to the people having it. What is underdeveloped is the machinery built to study it.
The instruments can't find what people plainly describe
The sharpest tension in the review sits between two kinds of evidence, and it is the same tension that runs through a great deal of neurodivergence research. The qualitative studies — the ones that asked autistic adults to describe their experience in their own words — were consistent and unambiguous. Participants described rejection in language like “misery,” “suffering,” and feeling “horribly rejected.” Across interviews and written accounts, the picture held together: rejection lands hard, and the distress is real and substantial.
The experimental studies pointed in no clear direction at all. Of the studies comparing autistic to non-autistic responses, three found greater distress, two found lower distress, two were mixed, and one found no difference. The lab paradigms — Cyberball, a virtual ball-tossing game rigged to exclude the participant, and the Social Judgement Task, which delivers fake social feedback — could not produce a stable answer to a question the interview studies answered cleanly.
The authors’ own reading of this is the important part, and it points at the instrument rather than the person. The experimental paradigms may simply not be sensitive enough to register the experience in autistic samples. Self-report measures of distress can be confounded by alexithymia — difficulty identifying and naming one’s own internal states — so an autistic participant may feel something intensely and report it flatly. A rigged ball game in a lab does not capture how rejection works in an autistic person’s actual life, where it accumulates across years of misread intentions and social exclusion. The measure returns noise not because nothing is there, but because the measure was built to detect something else, in someone else.
This is the same structural problem the ageing-in-autism literature ran into, where objective cognitive tests found normal ageing while autistic adults reported decline the tests could not see. When a careful person’s plain description of their own experience consistently disagrees with the instrument meant to measure it, the honest move is not to trust the instrument and discount the person. It is to ask whether the instrument can see the thing at all. The interview studies and the lab studies are not really in conflict. One captured the experience; the other used tools not built to.
Two stories about where it comes from — and why the difference matters
Underneath the measurement question sits a deeper one the review surfaces without resolving: where does this distress come from? Two accounts are on the table, and they carry very different implications. The original framing presents RSD as something close to intrinsic — a genetic, neurological feature of the brain, an innate hypersensitivity wired in, largely resistant to psychological intervention and responsive mainly to medication. The competing account, which the review’s evidence leans toward, locates the distress in cumulative social experience rather than innate wiring.
That second account has real weight behind it. A recent meta-analysis the review cites found autistic people face dramatically elevated lifetime rates of bullying, violence, emotional abuse, and adverse interpersonal encounters compared to non-autistic peers — bullying at 46 to 69% against 2 to 17%. The qualitative studies located the distress in exactly this terrain: being repeatedly misunderstood, having one’s differences met with intolerance, being ghosted on dating apps with no explanation, a pervasive lack of social safety. These resemble what the literature calls minority stressors — the accumulated weight of stigma and marginalisation — which are known to produce heightened vigilance to rejection and amplified responses to it. On this reading, an intense reaction to rejection is not a glitch in the autistic brain. It is what happens to anyone who has been rejected often enough, by a world that kept telling them they were wrong.
The difference between these two stories is not academic, because they point at different things to change. If the distress is intrinsic neurology, the response is to manage the individual — medication, regulation, adjusting the person to a world that stays as it is. If the distress is the accumulated residue of a hostile social environment, the response is to change the environment — to reduce the rejection, build the social safety, address the intolerance that produced the wound in the first place. The first story locates the problem in the autistic person. The second locates it in what has repeatedly been done to them. The review cannot say definitively which is right, and very likely both contribute, but it is honest about which way the current evidence tilts — and it tilts toward the environment.
This is where the research gap stops being merely an academic inconvenience and becomes something with stakes. As long as RSD remains under-studied and conceptually unsettled, the intrinsic-neurology framing — the one with the longest head start and the simplest clinical action — fills the vacuum by default. The more carefully the experience is actually investigated, on the community’s own terms and with autistic people involved in the research, the more the environmental account comes into view. Getting the science to catch up is therefore not just about validating an experience that is already real to the people having it. It is about which story gets told about why, and what that story asks the world to do.
For now, the honest position is the one the review holds: the experience is real and often profound, the science is fragmented and years behind, and the most consistent thread in the evidence points outward at the social world rather than inward at the autistic brain. That is a useful place to leave it heading into the summer — not a neat conclusion, but a clear direction of travel, and a reminder that on this subject the people living the experience got there first.
Citations
van Asselt, A., Reekers, D. & Roke, Y. (2026) — Rejection Sensitivity Dysphoria in Autistic Adults: A Scoping Review — Neurodiversity
Cooke, K., Ridgway, K., Pecora, L. et al. (2025) — Gender differences in the prevalence of autistic experiences of interpersonal violence: A mixed methods systematic review and meta-analysis — Review Journal of Autism and Developmental Disorders
Levi, M. K., Schreuer, N., Granovsky, Y. et al. (2023) — “Feeling unwanted, when nobody wants you around”: Perceptions of social pain among people with autism — The American Journal of Occupational Therapy
Eisenberger, N. I. (2012) — The pain of social disconnection: examining the shared neural underpinnings of physical and social pain — Nature Reviews Neuroscience
