Key moments, in order
Criterion A is persistent deficits in social communication and social interaction across multiple contexts, in three subcategories.
Social communication problems do not mean autism. As with almost every DSM disorder, multiple things can explain the same deficits.
Criterion B is restricted, repetitive patterns of behavior, interests or activities, two of four. The first is stereotyped or repetitive movements, use of objects or speech, such as hand flapping, lining up toys or echolalia.
It is a neurodevelopmental disorder: the features must be present in the early developmental period and come from neurodevelopment.
It is typically recognized in the second year of life. Regression after age two does not rule it out, but look more closely at other explanations.
We have seen numerous adults first evaluated and diagnosed with autism after the clear onset of schizophrenia.
The neurodiversity movement frames autistic traits as healthy human variation, but in some sense, rather than de-pathologizing autism, we have pathologized normal traits.
Criterion D requires clinically significant impairment, and E requires that it is not better explained by intellectual disability or global developmental delay.
The DSM is not aggressive enough about "not better explained by," probably because the people who framed the diagnosis mostly work with children.
Many patients with personality disorders feel a neurodevelopmental label fits them better, and a clinician should be able to tell when personality pathology is what is going on.
There is a large mutual blind spot between the clinicians and researchers who focus on personality disorders and those who focus on neurodevelopmental disorders, with too little coordination.
Overdiagnosed or underdiagnosed is a false dichotomy. Both are true.
Specifiers include with or without intellectual or language impairment, association with a known genetic, medical or environmental factor, and association with another neurodevelopmental, mental or behavioral problem.
Level 1 is "requiring support." That tells you something: a DSM-defined autism spectrum disorder means you need some kind of support, and that is probably where the bar for disorder sits.
Autism is a gigantic bucket of very different pathologies. The same diagnosis covers a top-tier programmer who needs prompting for appointments and a person whose main activities besides self-care were spinning and hitting people, and a category that broad tells us little.
Executive dysfunction is a big one. ADHD used not to be diagnosed alongside autism because executive dysfunction was considered part of autism, and it may truly be a separate issue from ADHD.
A typical defining feature is an uneven profile of abilities: incredibly intelligent in one area, with a mismatch in others.
Theory of mind problems are common, but their source and how they show up differ from the mentalization problems of personality disorders.
Also common: motor clumsiness, GI problems, sleep-wake and circadian problems, self-injury under distress, emotional dysregulation and irritability.
Childhood-onset schizophrenia has a period of normal or near-normal development before the prodrome, and the prodrome's atypical interests and social impairment can look like autism.
Autism and schizophrenia can co-occur, probably with a shared genetic risk pool, and it can get very confusing which one you are dealing with at a given time.
The DSM notes that autistic behavior can be perceived as narcissistic, schizotypal or schizoid personality, and the opposite is true too.
If someone goes 30 years in a modern Western society without an autism diagnosis and then develops new difficulties, consider everything else that could be interacting with a baseline of autistic traits.
No single theory can explain something as complex as the brain. Treat the theories as complementary snapshots: sometimes an older one captures a patient better, sometimes a newer one.
The social motivation hypothesis says people with autism find social input less rewarding and pay less attention to faces and gaze.
We do not buy its assumption that social capacity is simply learned; it is something more fundamental in the biology of brain functioning.
Weak central coherence theory and the mind-blindness theory of theory of mind deficits are two other frames.
The most compelling one, though not the most politically correct, is systemizing versus empathizing, with the general population spread along both.
Systemizing is the drive to analyze or build systems: picking up the rules and regularities that govern a system so you can predict how it will behave.
Split empathizing into cognitive empathy, intellectually understanding what someone feels or wants, which is learnable, and affective empathy.
People with clinically identifiable autism show more systemizing than even the general-population male, whatever their own sex.
There may also be a double empathy problem, in which other people fail to empathize with the autistic person.
Some workplaces, such as Microsoft and Ernst & Young, have tried management styles, rules and environments suited to people with autistic traits.
Failure of empathy does not explain borderline conditions; it is not a well-supported hypothesis when empathizing is measured in people with BPD.
Psychopaths have the opposite profile: strong cognitive empathy and poor affective empathy. Autistic people appear to have some internal affective reaction to others.
People with BPD are better at detecting negative reactions, interpersonal threat and disapproval, and being so good at that ruins their ability to read other mind states accurately.
Autism spectrum disorder is high systemizing and low empathizing, plus a collection of associated problems (repetition, emotional regulation problems, some learning problems) downstream of genetics and neurobiology, at a level that makes it a disorder rather than traits.
Each moment is a point made in the episode, linked to the second it was said. The topic links open everything the podcast has said on that topic.