Understanding Autism: The Spectrum, the Levels and the Diagnosis

Understanding Autism: The Spectrum, the Levels and the Diagnosis

There are around 290,900 autistic Australians — about 1.1% of the population, up 41.8% from 205,200 in 2018.

That increase is the single most misread statistic in this field, so it is worth dealing with first. It is not an epidemic. It reflects broader diagnostic criteria, better recognition — particularly in girls and women — greater awareness, and the strong incentive the NDIS created to obtain a diagnosis in order to access support. The underlying rate of autistic people in the population is not thought to have changed much. What has changed is how many of them get identified.

This guide covers what autism is, what the diagnostic levels mean and why many autistic people object to them, what the Australian data shows, which conditions commonly occur alongside autism, and how diagnosis works here.

It is written for autistic people, families and support workers. It is general information, not clinical advice.

A note on language before we start: this guide mostly uses identity-first language — “autistic person” rather than “person with autism” — because that is the strong preference of most autistic adults and advocacy organisations in Australia. Some families prefer person-first language, and that is their call. Follow the individual.

What autism is

Autism is a neurodevelopmental difference affecting how a person communicates, processes sensory information, relates socially, and manages change and routine. It is lifelong, present from early development, and not caused by parenting, vaccines or anything a family did.

Under the DSM-5 — the diagnostic manual used in Australia — a diagnosis requires differences in two domains:

Persistent differences in social communication and social interaction, which may include differences in social-emotional reciprocity, in non-verbal communication, and in developing and maintaining relationships.

Restricted, repetitive patterns of behaviour, interests or activities, which may include repetitive movements or speech, insistence on sameness and routine, highly focused interests, and differences in sensory processing — either heightened or reduced sensitivity.

Those are the clinical criteria. They are written in deficit language, which is worth noticing, because the same features described differently are: direct communication, deep focus, pattern recognition, consistency, strong memory for detail, and heightened sensory awareness.

Both descriptions are of the same thing. Which one gets used shapes how an autistic person is treated.

What autism is not

Each of these costs people something.

It is not a disease or an illness, and it is not something to be cured. Autistic-led organisations are generally emphatic about this, and it is worth taking seriously rather than treating as semantics. A support approach aimed at making someone appear non-autistic has different outcomes from one aimed at helping them function and be understood — and the research on masking suggests the first approach does measurable harm.

It is not an intellectual disability, though the two co-occur. Many autistic people have average or above-average intelligence; some have intellectual disability. The two are separate.

It is not a behaviour problem. Distress, meltdown and shutdown are responses to overwhelm, unmet need, pain or sensory load — not misbehaviour. Our sensory processing guide covers the distinction.

It is not caused by vaccines. This has been investigated exhaustively and comprehensively disproven.

It does not look one way. The stereotype — a young boy, non-speaking, with obvious repetitive movements — describes a minority. Autistic people include talkative extroverts, people in senior professional roles, parents, and people who were diagnosed at 50.

It is not the same as being “a bit autistic.” Autism is not a personality quirk everyone has to some degree. It is a diagnosable neurodevelopmental difference with substantial functional consequences for most autistic people — the ABS found 82.4% had a limitation in at least one core activity.

The levels, and why they are contested

The DSM-5 describes three levels of support need:

Two things need saying about these, and they conflict.

Administratively, the levels matter enormously in Australia. Autism at Level 2 or 3 appears on NDIS List A; Level 1 does not. The level written on a diagnostic report determines which access pathway applies. Our autism and the NDIS guide covers this in detail, because it is the practical fact most families are never told.

Clinically and politically, the levels are weak. The objections are substantive:

Support needs fluctuate. The same autistic person may need Level 1 support in a familiar, predictable environment and Level 3 support in an unfamiliar one, or during burnout. A single number describes neither.

The levels collapse different domains into one. Someone may need very substantial support with communication and minimal support with daily living, or the reverse. The levels force that into one figure.

They describe support needs, not the person. A level is a statement about environment and support as much as about the individual — which means it changes when support changes.

They are used as functioning labels. “High functioning” and “low functioning” are widely rejected by autistic people, because they flatten a multi-dimensional picture and are usually wrong in both directions. Someone described as high functioning may be unable to make a phone call; someone described as low functioning may have intact comprehension and no way to demonstrate it.

The practical position: treat the level as an administrative requirement to get right on paper, and as a poor guide to what an individual actually needs. The useful information is in a functional assessment and in what the person says about their own life.

What the Australian data shows

From the ABS Survey of Disability, Ageing and Carers 2022:

Population. 290,900 autistic Australians, or 1.1%, up 41.8% from 205,200 (0.8%) in 2018.

Sex. Prevalence was 1.6% for males and 0.7% for females — roughly twice as common in males, down from more than three times as common in 2018. That narrowing is one of the more significant findings in the dataset. It is not a change in who is autistic; it is improved recognition of autistic girls and women, who were historically missed at a high rate because diagnostic criteria and clinical expectations were built around male presentation.

Age. Prevalence peaks in the 10–14 age group and declines from the late teens: 3.1% among those under 25, and 0.3% among those 25 and over. Autism does not resolve at 25. That gap is under-identification in adults — people who grew up before current diagnostic practice. Data for people aged 40 and over was not published because estimates were unreliable, which tells its own story.

Severity. About 73% of autistic people had a profound or severe disability, and 82.4% had a limitation in at least one core activity. Profound or severe restrictions were reported for communication (59.7%), mobility (60.4%) and self-care (47.7%).

Education. 45.1% of autistic people aged 15 and over had completed Year 12. Only 5.2% held a bachelor’s degree or higher, against 35.3% of people without disability.

Employment. Among working-age autistic people, 50.2% were in the labour force and 42.2% were employed — up substantially from 27.3% in 2018 — with an unemployment rate of 18.2%. The improvement is real and the gap remains large.

Need for assistance. 84.0% needed help with at least one activity, and 59.2% needed it daily. Of those needing help, 66.4% reported unmet need, up from 51.7% in 2018.

That last figure deserves emphasis. Unmet need among autistic Australians rose by nearly fifteen percentage points in four years, during a period when the NDIS was growing rapidly. Whatever else is true about autism support in Australia, it is not that need is being met.

In the NDIS, autism is the largest primary disability group by a wide margin: 338,099 participants as at 31 March 2026, or 44% of all 774,456 participants.

Co-occurring conditions

This is where a great deal of the functional impact actually sits, and where support plans most often fall short — because they are written about autism and ignore everything alongside it.

Commonly co-occurring:

ADHD, very frequently. The combination is common enough that many clinicians now assess for both routinely, and it changes what support helps — executive function difficulties compound.

Anxiety disorders, at substantially elevated rates.

Depression, also elevated, and frequently related to social exclusion, masking and unmet need rather than to autism itself.

Intellectual disability, in a significant minority.

Epilepsy, at rates well above the general population.

Sleep disorders, extremely common and frequently untreated. Poor sleep worsens everything else — sensory tolerance, emotional regulation, executive function — so it is one of the highest-value things to address.

Gastrointestinal conditions, including constipation and reflux, common and frequently missed, particularly in people who cannot easily report pain.

Feeding and eating differences, including highly restricted diets driven by sensory factors, and avoidant/restrictive food intake disorder.

Dyspraxia and motor coordination difficulties.

Hypermobility and related connective tissue conditions, increasingly recognised as co-occurring.

Alexithymia — difficulty identifying and describing one’s own emotional states — which affects a substantial proportion of autistic people and has consequences for mental health care and for reporting pain.

PTSD and trauma responses, at elevated rates. Autistic people experience bullying, exclusion and abuse at higher rates than the general population.

The practical point for anyone writing or reading a support plan: autism alone is frequently not the main driver of support need. A plan that addresses autism and ignores co-occurring ADHD, chronic sleep deprivation, untreated anxiety and constipation will underperform regardless of how well-written its autism content is.

Monotropism and the attention theory

Worth including because it is the single most useful explanatory framework for autism that most people have never heard of, and it was developed substantially by autistic researchers.

Monotropism proposes that autistic attention tends to be allocated to fewer interests at a time, more intensely — a narrow, deep attention tunnel rather than a broad, shallow one. Non-autistic attention is described as more polytropic: spread across more things, less deeply.

If that is right, a great deal follows naturally from one mechanism:

Deep focus and special interests are the same phenomenon as difficulty attending to several things at once.

Difficulty with transitions makes sense, because leaving an attention tunnel is costly, not merely inconvenient. Being interrupted mid-task is not a minor irritation; it is losing something that took effort to enter.

Social difficulty follows partly from the fact that conversation requires rapid parallel attention — words, tone, expression, body language, context, one’s own response — which is exactly what a monotropic attention style handles badly.

Sensory overload follows from reduced capacity to filter competing inputs.

Flow states and the distress of being pulled out of them are both explained.

Inertia — difficulty starting and difficulty stopping — becomes one phenomenon rather than two contradictory ones.

The practical value is in what it suggests. If transitions are costly, warning and preparation help. If attention is a tunnel, then interruption is expensive and uninterrupted time is a support. If conversation is parallel-processing-heavy, then written communication and processing time are accommodations rather than concessions.

Monotropism is a theory rather than settled fact, and not everything about autism reduces to it. But it is a better predictive model than most, and autistic people frequently describe it as the first explanation that matched their experience from the inside.

How diagnosis works in Australia

For children, assessment is usually multidisciplinary — a paediatrician or psychiatrist alongside a psychologist, speech pathologist and occupational therapist. It involves developmental history, standardised assessment tools, direct observation, and information from parents and often school.

Waiting lists in the public system are long — commonly many months, sometimes over a year. Private assessment is faster and costs substantially more.

For adults, assessment is harder to obtain. Fewer clinicians are experienced in adult autism, waiting lists are long, private assessment commonly costs well over a thousand dollars, and Medicare rebates are limited. Assessment involves current functioning plus developmental history, which can be difficult to establish when records don’t exist and parents aren’t available.

What a diagnostic report should contain, from an Australian practical standpoint:

• Confirmation of the diagnosis against DSM-5 criteria
• The level, explicitly stated — this is the item most often missing, and it determines the NDIS access pathway
• Description of functional impact across domains
• Co-occurring conditions identified or flagged for assessment
• Recommendations

If a report omits the level and the clinician considers Level 2 or 3 appropriate, ask for an addendum. It is usually a short letter and it changes which NDIS pathway applies.

On whether to seek diagnosis at all — a real question for adults in particular. Reasons people do: access to support and funding, workplace and educational adjustments, self-understanding, and community. Reasons people hesitate: cost, the risk of being treated differently, concerns about travel or immigration in some countries, and the reasonable view that self-knowledge doesn’t require a clinician’s sign-off. Both positions are defensible. Diagnosis is necessary for NDIS access and not necessary for being autistic.

What the research is and isn’t working on

Briefly, because families newly navigating a diagnosis get told contradictory things.

Causes. Autism is substantially heritable, with many genes involved, each contributing small effects, alongside some environmental factors operating prenatally. There is no single cause and no single gene. It is not caused by parenting or vaccines.

Early support. Evidence favours early, naturalistic, child-led approaches that work with autistic development rather than against it, and that involve families. Approaches aimed at suppressing autistic behaviours to make a child appear non-autistic are increasingly criticised by autistic adults and researchers, including on the grounds that masking is associated with poorer mental health outcomes.

Autistic-led research is a growing and important shift. Much historical autism research was conducted without autistic involvement and asked questions autistic people didn’t consider priorities. Participatory research has changed what gets studied — autistic burnout, masking, and the double empathy problem all emerged substantially from autistic-led work.

What there isn’t: a cure, and no serious researcher is looking for one. Claims of recovery, biomedical cures, chelation, restrictive diets presented as treatment, or similar should be treated with scepticism and discussed with a doctor before trying. Some are expensive; some are dangerous.

Language, and why autistic people care about it

Worth a section, because getting it wrong undermines everything else you say.

“Autistic person” is preferred by most autistic adults and advocacy organisations. “Person with autism” is preferred by some families and some clinicians. The convention is shifting toward the former. Follow the individual.

“ASD” and “autism spectrum disorder” are the clinical terms and are accurate in clinical contexts. Many autistic people dislike “disorder” and simply say “autism” or “autistic.”

Avoid “high functioning” and “low functioning.” Widely rejected, for the reasons set out above.

Avoid “suffers from autism,” “afflicted with,” and “person living with autism.”

“Neurodivergent” describes any neurological difference from the typical, including autism, ADHD and others. “Neurotypical” describes people without those differences. “Neurodiversity” describes the fact of variation across a population — it is not a synonym for autism, and an individual is not “neurodiverse.”

Avoid describing autism as a superpower, which autistic people also generally dislike. It sets an expectation of exceptional ability, and most autistic people are ordinary people who need support, like anyone else.

Don’t say “you don’t seem autistic.” It is usually intended as a compliment and lands as a dismissal of effort, often substantial effort. Our guide on masking and burnout covers why.

What the label tells you, and what it doesn’t

A closing point that matters more than any figure above.

Knowing someone is autistic tells you that they process social information, sensory input, routine and change differently from most people. It tells you very little else — and specifically, it tells you nothing reliable about:

Whether they will speak. Some autistic people are non-speaking, some have fluent speech, many are somewhere in between or lose speech under stress.

What they will understand. Comprehension and expression are frequently very different, and assuming the first from the second is one of the most common and most damaging errors made about autistic people.

What they will achieve. Autistic Australians work, study, parent, run businesses and hold public office — and the employment rate rose from 27.3% to 42.2% in four years, which suggests the ceiling was always about environments rather than people.

What support they will need. That varies by person, by environment, by day, and by whether they are in burnout.

The practical consequence is that the diagnosis is a starting point for questions rather than an answer. The useful information is in a functional assessment, in what the person says about their own experience, and in what they want to be doing.

Where to next

Autism and the NDIS covers access, including the List A distinction. Thriving Kids and foundational supports covers the reform affecting children aged 8 and under. Autistic adults, masking and burnout covers late diagnosis and the adult picture the data barely captures.

We’re a registered NDIS provider supporting autistic adults across NSW, in supported independent living homes, SDA dwellings and people’s own homes. Call 1300 271 824.