At UCLA’s PEERS Clinic, I coached autistic teenagers learning social skills that could help them navigate adulthood. Later, as a behavioral therapist, I worked directly with autistic individuals and their families in their homes.
One mother stopped me after a session. Her daughter had just turned eighteen, and she was trying to understand what came next.
Her daughter had spent years inside a system that knew her: teachers who understood how she communicated, individualized goals, adults who knew that her going quiet did not mean she had stopped paying attention. Her mother had spent those same years learning to advocate inside it: which forms mattered, which meetings to attend, how to ask.
Nothing about her daughter’s needs had changed because she turned eighteen. The system around her had.
I used to think of this as an autism problem. After working with older adults, I no longer do.
The United States has built sophisticated care around particular stages of life and particular diseases. We are much worse at the space between them.
For autistic people, one of the sharpest transitions comes at the end of school. It’s often described as services ending at eighteen, but that shorthand hides what actually happens. Under the Individuals with Disabilities Education Act, eligible students hold an entitlement, and an individualized team is responsible for coordinating supports, accommodations, and transition planning. When a student leaves that system, adult support becomes largely eligibility-based: Medicaid waivers, developmental disability agencies, vocational rehabilitation, housing, transportation. Each has its own application, its own proof of need, its own queue.
The person has not become harder to support. Responsibility has migrated, from an institution to an individual and, often, their family.
And eligibility is not access. In 2025, more than 600,000 people were on waiting or interest lists for Medicaid home- and community-based services across 41 states, most of them people with intellectual or developmental disabilities. KFF cautions that these lists imperfectly measure unmet need, because states differ in how they determine eligibility. They are still evidence of how difficult timely access can be.
Research documents the transition more directly. A longitudinal study followed 204 autistic people across 14 years surrounding high-school exit. Service use declined afterward, and unmet needs rose immediately, for autistic people both with and without intellectual disability. The authors tied those patterns to the loss of entitlement to school-based services and the limited availability of adult ones.
Years later, I started seeing another version of this while working with older adults.
At first the problems looked small. One older woman slid her phone toward me because she could not get into her patient portal. Through AgeWell, we heard repeatedly from older adults and caregivers about the difficulty of managing medications, and built multilingual reminders in response.
None of these were failures of medical knowledge. These people often had excellent physicians. What was failing was everything between the appointments.
A physician can prescribe the right medication, but someone has to remember when to take it. A hospital can discharge a patient with instructions, but someone has to understand them, reconcile the medications, book the appointments, and notice when something is going wrong.
Often, that someone is family. More than 51 million Americans provide care to someone age 50 or older, spending an average of 26 hours a week doing it. We call them daughters and husbands and sons. In practice they also become schedulers, medication managers, drivers, medical historians, insurance navigators, and interpreters of subtle change. None of those jobs appears on an insurance card. Remove that person, and a carefully designed care plan can begin to come apart.
Which brings me to the part I can’t stop thinking about. Imagine an autistic man who is 55. His parents are 80.
For four decades they have known that when he stops eating a particular food, something may be wrong. They know which rooms overwhelm him, which routines settle him, how he shows pain, which medications he has tried, which clinicians listen and which don’t, and which forms have to be filed each year to keep his supports in place.
That knowledge may exist largely in two people’s heads. And his parents are developing health problems of their own.
Who inherits forty years of understanding?
We do not have a good answer. Autism research has concentrated heavily on childhood; one review found that work involving older autistic people accounted for only about 0.4% of published autism studies over the decade it examined. So we know remarkably little about what happens when lifelong communication and sensory differences meet dementia, polypharmacy, reduced mobility, and the loss of the people who have navigated care for decades.
An autistic adult can therefore meet related forms of fragmentation twice in one life: once leaving childhood systems, and again aging into increasingly complex needs. These are usually treated as separate policy problems. I think they expose the same weakness.
We have built care around institutions with bounded responsibilities rather than around continuity across a person’s life. Schools are responsible until they are not. Pediatric clinicians hand off.
Specialists treat the systems assigned to them. Families fill much of the space left between. What if continuity were the thing we funded?
For young adults leaving school, that could mean transition coordinators who remain involved after graduation, bridging healthcare, benefits, employment, and community services instead of handing a family a list of phone numbers.
For adult medicine, it means stronger autism education in residency and continuing medical education: communication accommodations, sensory needs, supported decision-making, and the danger of dismissing every new symptom or behavioral change as “just autism.”
For older adults, it means treating care coordination and caregiver support as central components of healthcare rather than social afterthoughts.
None of this assumes autistic people need lifelong supervision. Independence should be supported wherever the person wants it. But independence is not the same as being left alone. Someone can direct their own life and still need help understanding a scan, reaching a clinic, or making a decision heard.
Support and autonomy are not opposites. Often, support is what makes autonomy possible.
I want to know whether a treatment exists. I also want to know whether the person can reach it, understand it, afford it, manage it, and keep receiving it when the institution around them changes.
Because an autistic child does not become a different person when school ends. An older adult does not become a list of diagnoses because five specialists are involved. And an autistic adult does not stop being autistic when they grow old.
The paperwork changes. The person remains.
Aishwarya Manga is a former behavioral therapist and social skills coach at UCLA’s PEERS Clinic, and the founder of AgeWell Bruins, an intergenerational senior health organization.
