For decades, autism research has focused almost entirely on children, early intervention, and education. Now, a rapidly growing population of older autistic adults is exposing a major blind spot in global health science: no one fully knows what ageing with autism looks like. Reporting highlighted in Nature reveals that while awareness and diagnoses have surged, research into later life remains scarce, leaving millions navigating ageing without evidence-based care.
Sondra Williams spent much of her youth and early adulthood moving through mental-health systems, treated for conditions ranging from schizophrenia to severe anxiety. It was not until her late 30s, after two of her children were diagnosed with autism, that clinicians recognized she met the criteria as well. The diagnosis reframed her life story, transforming years of confusion into clarity. Now 63, Williams faces type 2 diabetes and high blood pressure, conditions some studies suggest may occur more frequently among autistic people, and she worries about what lies ahead.
Her experience reflects a demographic shift unfolding worldwide. An analysis of data from more than 200 countries found that the number of autistic people aged 70 or older increased from fewer than one million in 1990 to nearly 2.5 million in 2021. That figure could exceed five million by 2040. Population growth explains part of the rise, but improved awareness, expanded screening, and updated diagnostic definitions have also allowed adults to be identified for the first time.
Yet despite this surge, researchers know remarkably little about autism in midlife and old age. Since 2012, only a tiny fraction of autism studies have included older participants. Historically, autism science centered on early development, while gerontology studies often excluded neurodivergent individuals altogether. The result is what some experts call a “lost generation” of autistic adults who grew up without recognition, support, or inclusion in research.
Autism itself was only identified as a distinct condition in the 1940s and was long viewed as related to schizophrenia. Not until 1980 did diagnostic manuals formally recognize it as a separate condition, and the broader concept of autism spectrum disorder, including adult diagnosis, was introduced as recently as 2013. Because of those shifting definitions, large numbers of adults remain undiagnosed. Surveys in England suggest that most autistic people over 50 may never have received a formal diagnosis, complicating efforts to study their health over time.
The lack of data is striking when compared with research on younger populations. Tens of thousands of studies have examined autism in childhood, but only a few hundred have explored ageing. Gregory Wallace, a developmental neuropsychologist at George Washington University, summarizes the situation bluntly: scientists are still trying to establish the most basic understanding of what ageing means for this group.
What evidence does exist suggests autistic adults may face elevated risks for several medical conditions. Analyses of health records from thousands of older adults show higher rates of mood disorders, epilepsy, gastrointestinal problems, osteoporosis, and cardiovascular disease compared with non-autistic peers. Some datasets also indicate increased prevalence of Parkinson’s disease and dementia. A study examining tens of millions of medical records found that as many as 35 percent of autistic adults over 64 had a dementia diagnosis, far higher than in the general population.
Researchers caution that interpreting these numbers is complex. Zheng Wang, a neuroscientist at the University of Florida, is investigating whether movement-related differences observed in autistic adults stem from neurodevelopmental traits present throughout life or from neurodegenerative changes that emerge with age. Untangling lifelong variation from disease processes remains one of the field’s greatest challenges.
Medication history may also play a role. Many autistic people were prescribed antipsychotics, anti-seizure drugs, or other psychotropic medications earlier in life, sometimes for decades. Long-term use of such drugs can produce side effects that resemble age-related neurological conditions, potentially clouding diagnoses later on.
Still, the picture is not entirely one of vulnerability. Some studies suggest that autistic brains may age differently, and in certain respects more slowly. Hilde Geurts, a neuropsychologist at the University of Amsterdam, has found that differences between autistic and non-autistic adults do not necessarily widen with age, challenging assumptions about accelerated decline.
Similarly, Blair Braden at Arizona State University has used brain imaging to track memory and structural changes over time. Her work shows a highly varied pattern: some autistic adults experience faster declines in long-term memory, while others remain stable or even demonstrate strong compensatory strategies. This variability underscores a central feature of autism itself: it is not a single pathway but a spectrum shaped by biology, environment, and lived experience.
That diversity makes research especially difficult, says Ruth Carper, an imaging neuroscientist at San Diego State University. By the time individuals reach later life, decades of distinct personal histories interact with neurological differences, producing outcomes too varied for simple conclusions.
Another underexplored dimension is gender. Autism was long assumed to be far more common in boys, leading to a shortage of studies on women and girls. Researchers are now examining how ageing intersects with reproductive health, menopause, and hormonal changes. Early findings suggest autistic women may experience more intense psychological and physical symptoms during menopause, including anxiety, fatigue, and pain, yet often receive little tailored medical guidance.
To close these gaps, scientists are launching long-term studies that follow autistic adults across years rather than capturing a single snapshot. Lauren Bishop, a social scientist at University of Wisconsin–Madison, is leading one such effort tracking adults over 40 to identify markers of cardiovascular, cognitive, and metabolic ageing. These longitudinal projects are expected to provide the first comprehensive datasets on how autism interacts with the biology of ageing.
Advocates stress that research must include the voices of autistic people themselves. For many, the issue is not simply scientific curiosity but quality of life. Better understanding could lead to earlier screening for chronic illnesses, improved mental-health care, and social supports tailored to sensory and communication needs in older age.
Gavin Stewart, a developmental psychologist at King’s College London, describes the field as still “setting the scene,” with practical applications only beginning to emerge. The urgency, however, is growing alongside the population it seeks to serve.
For Williams, the stakes are deeply personal. She hopes future research will spare younger generations of autistic people from the uncertainty she now faces. As awareness expands, scientists are racing to answer a question that should have been asked long ago: not just how autistic people develop, but how they grow old.
