Half a million young people, and a split that runs in opposite directions
Lovisa Hellsten and colleagues at Karolinska Institutet, working with Region Stockholm and collaborators at Drexel and Aarhus, have published a population-based registry study in Autism Research covering every 12- to 22-year-old resident in Stockholm County between 2018 and 2022 — 454,405 people, of whom 11,850 were autistic, 31,144 had ADHD, and 9,126 had both. Sweden’s youth clinics provide free, confidential sexual and reproductive health care to this age group, and the study tracked 492,762 consultations across the period, in person and digital.
The headline split runs in opposite directions, which is the first thing worth sitting with, because autism and ADHD are routinely bundled together in service planning. Young people with ADHD used these services considerably more than their peers — with adjusted odds of an in-person visit roughly double for females and around half again as high for males. Autistic young people used them considerably less: adjusted odds of an in-person visit around a third lower for females, and around half as likely for males, the lowest uptake of any group in the study.
Set against what these services do, the autistic figure is the one to worry about. Youth clinics provide contraceptive counselling, sexually transmitted infection testing, emergency contraception and psychosocial support — and the population using them least is one with documented elevated rates of sexual victimisation, frequently poor access to appropriate sex education, and, among autistic women, higher rates of menstrual and hormonal difficulty. Lower use is not lower need. It is need meeting a service that is somehow not reachable, and the study’s real contribution is in locating where the failure occurs.
An intensity that matches, once the threshold has been crossed
The locating finding is this. When the researchers looked past whether young people used the service at all, and measured how intensively those who did use it — the rate of visits across the study period — autistic females were comparable to their non-diagnosed peers. Fewer of them got there. Those who did behaved like everyone else.
That distinction does a great deal of work. It rules out the explanation that would otherwise be reached for first: that autistic young people are less interested in relationships or sexual health, or disengage from care once involved. If disengagement were the mechanism, intensity would be lower too. It isn’t. What is lower is the probability of crossing the threshold in the first place — which points at everything that happens before the appointment: booking, phoning, the waiting room, the unfamiliar building, the noise, the uncertainty about what will be asked.
The pattern of who autistic young women saw once inside sharpens this further. They were more likely than their peers to have seen a physician, and substantially less likely to have seen a midwife — despite midwives providing the overwhelming majority of care at these clinics. And they were markedly less likely to have been dispensed short-acting contraception, while being no less likely than their peers to have received a long-acting method. So the picture is not a population avoiding sexual health care wholesale. It is a population arriving through different routes, at different points, for different reasons — plausibly, as the authors suggest, for the more complex issues that reach a physician, including menstrual and hormonal difficulty rather than contraception alone.
Autistic males are the group the study should worry anyone most. They had the lowest uptake of any group, with either provider, and their intensity of use was persistently lower too — so for them, unlike for autistic females, the difficulty does not stop at the door. That is a group essentially outside preventive sexual health care, and the study offers no comfortable account of why.
When changing the door changes who comes through it
The most actionable finding concerns mode of contact. Digital visits — video and chat consultations, expanded since 2020 — were a small minority of all contacts, and the study is honest that this is an early phase of implementation. But the gap between autistic young people and their peers was significantly narrower for digital contacts than for in-person ones, for both females and males.
The same young person, the same service, the same clinical content. Change the route in — remove the building, the waiting room, the sensory load, the face-to-face social demand, the requirement to speak aloud about something difficult — and the disparity shrinks. Nothing about the young person changed. The door changed.
This is the demand-structure argument arriving as health-service data, and it is unusually clean because the comparison is internal. The barrier cannot be located in autistic characteristics, because the autistic characteristics are constant across both modes and only the access route varies. Whatever is suppressing uptake lives in the in-person pathway — its sensory environment, its communication demands, its unpredictability — and a route that removes those brings people closer to parity. The authors are careful to note that it remains unclear whether digital access improves outcomes or simply increases the volume of contacts, and that caution is right. But as a signal about where the obstruction sits, it is hard to misread.
There is a caution worth carrying alongside it. The finding is not that digital replaces in-person care; contraceptive insertion, examination and testing require a room and a clinician. It is that digital lowers the threshold for the first contact, after which the study suggests engagement holds up. Hybrid pathways — digital entry, in-person continuation — are the obvious implication, and the authors note these remain underexplored.
A limitation deserves naming, because the authors name it themselves and it matters for how far the finding travels. The analysis uses legal sex only, and gender-diverse identities are more common among autistic people than in the general population, with their own well-documented barriers to sexual and reproductive care. There is no data here on co-occurring intellectual disability, none on the hormonal and gynaecological conditions that would help explain the physician-versus-midwife pattern, and diagnoses come from ICD-10 codes without any measure of severity. Sweden is also a digitally mature country with universal coverage — the digital finding may not transfer to systems without that infrastructure. And because undiagnosed autistic and ADHD people sit within the comparison group, the differences reported are likely to be underestimates rather than overestimates.
The explanation the data quietly withdraws
The ADHD result deserves its own attention, because it removes an explanation the literature has leaned on for years.
Elevated rates of teenage pregnancy, sexually transmitted infection and related outcomes among young people with ADHD are well documented. The standard account attributes them to impulsivity combined with insufficient or delayed preventive care — under-use of services as a driver of risk. This cohort used more: more visits with midwives and physicians alike, higher contraceptive dispensation overall, and notably higher uptake of long-acting reversible methods, with young women with ADHD nearly twice as likely as their peers to have been dispensed one. Rates of abortion — reported here as a service-use measure — were also higher.
Whatever explains the elevated outcomes, it is not that this group is failing to reach preventive services. They are reaching them at higher rates than anyone. The authors are appropriately careful about what replaces the discarded explanation, noting that dispensation is not the same as consistent use, and that switching or discontinuation of short-acting methods may matter more than whether a prescription was collected. That is a substantially different problem from under-attendance, and it points somewhere else entirely — at whether the method offered fits the person’s actual executive landscape, at whether follow-up exists, at what happens between the prescription and the daily practice. It is the sort of question that only becomes askable once the easier explanation is off the table.
Which is, in the end, what this study does across both populations: it moves the question from what these young people are failing to do, to what the service is failing to reach them with — a door that is too costly to walk through in one case, and a form of provision that stops at the pharmacy counter in the other.
Citations
Hellsten, L., Rast, J., Nielsen, A. et al. (2026) — Sexual and Reproductive Healthcare Among Youth With Autism and ADHD: A Population-Based Study From Sweden — Autism Research
Ames, J. L., Massolo, M. L., Davignon, M. N. et al. (2021) — Healthcare service utilization and cost among transition-age youth with autism spectrum disorder and other special healthcare needs — Autism
de Visser, R. O., Mosely, R., Gamble-Turner, J. et al. (2025) — Unmet need for autism-aware care for gynaecological, menstrual and sexual wellbeing — Autism
Skoglund, C., Kopp Kallner, H., Skalkidou, A. et al. (2019) — Association of attention-deficit/hyperactivity disorder with teenage birth among women and girls in Sweden — JAMA Network Open
