We Already Know What Causes Autism

For decades, the question of what causes autism has captured public attention, fueled debates, and spurred waves of research. Secretary of Health and Human Services Robert Kennedy has been a prominent and controversial figure in this space, firmly believing that environmental factors—such as dietary components, toxins, or vaccines—may play a significant role in the increased incidence and prevalence of autism. His advocacy has drawn both support and criticism, often putting him at odds with the scientific mainstream. After more than 30 years of research, the broad consensus is that autism is primarily biological in origin, rooted in genetics, and shaped by prenatal and early developmental factors. What has changed dramatically is not the disorder itself but how we define, detect, and respond to it.

The Illusion of an Epidemic

Autism, or autism spectrum disorder (ASD), was once considered a rare, profoundly impairing childhood condition. In the 1970s, autism was thought to affect approximately 1 in 2,500 children (Fombonne, 2003). Today, estimates place the prevalence at around 1 in 36 children (Maenner et al., 2023). The instinctive assumption—that something in the environment must be causing a dramatic rise—is misleading. The increase is not in actual incidence but in diagnosis. This is due to broader diagnostic criteria, better screening tools, increased awareness, and a shift in the kinds of children being labeled.

In the past, a diagnosis of autism was reserved for children with profound impairments in language, cognition, and social functioning. These children often presented with intellectual disabilities, self-injurious behaviors, or a complete lack of speech. The criteria were narrow, focusing on severe, unmistakable cases. Over time, the diagnostic definition evolved—particularly with changes in the Diagnostic and Statistical Manual of Mental Disorders (DSM). DSM-III (1980) included autism as a distinct disorder. Still, it wasn’t until the DSM-IV (1994) and the DSM-5 (2013) that the spectrum model solidified, expanding the boundaries of who could be diagnosed.

In our widely used textbooks Assessment of Autism Spectrum Disorder (Goldstein, Ozonoff & Naglieri, 2008; Goldstein & Ozonoff, 2018), we have documented this shift. Today’s children differ significantly from those diagnosed 30 or 40 years ago. Most have average or above-average intelligence, many are verbal, and a growing number are diagnosed not due to severe developmental delays but rather because of social communication differences and behavioral rigidity. Autism has shifted from being about profound disability to being more about differences in social learning.

This reframing is crucial. Autism today encompasses a range of children who may have previously been viewed as shy, quirky, or socially awkward rather than disabled. These children often attend regular schools, engage in conversations, and even excel academically—yet their social development follows a different trajectory. In no other area of child mental health has the diagnostic landscape shifted so dramatically in such a short time.

What Science Actually Tells Us

Research from the 1990s onward confirms the biological underpinnings of autism. Twin studies have consistently shown that autism is highly heritable, with concordance rates for identical twins ranging from 60 to 90 percent, compared to 0 to 30 percent for fraternal twins (Bailey et al., 1995; Hallmayer et al., 2011). More recent genetic studies have identified hundreds of genes associated with an increased risk for ASD, many of which are involved in brain development and synaptic functioning (Sanders et al., 2015).

Environmental factors do play a role—but not in the way that sensational headlines might suggest. Factors such as advanced parental age, prenatal exposure to certain medications, or complications during birth can slightly increase the risk. Still, they are not causes in a deterministic sense. They influence a child’s neurodevelopment in interaction with genetic susceptibility.

Vaccines, long a focal point of misinformation and controversy, have been definitively ruled out. Studies have shown no link between vaccination and autism (Taylor et al., 2014). The persistence of this myth has diverted attention from the real, evidence-based causes.

A New Understanding of Autism

Notably, the number of children with severe, impairing autism has remained relatively stable over the decades (Hyman et al., 2020). What has changed is the inclusion of children with milder symptoms under the same diagnostic umbrella. This contributes to the perception of an autism “epidemic” when, in fact, what we’re seeing is a diagnostic shift.

This shift reflects a broader trend in how we conceptualize child mental health. Autism is increasingly viewed as a neurodevelopmental difference rather than a disease. The focus has moved from “curing” autism to supporting neurodiverse individuals in navigating a world not built for them. Social skills training, behavioral therapies, and educational accommodations aim to help children with autism thrive on their terms, not to erase their differences.

Autism Essential Reads

A Brief History of the Science of Autism Spectrum Disorder

Autism, Inflammation, and the Gut-Brain Axis

Understanding that autism is a biologically based condition with a wide range of presentations doesn’t minimize the challenges many individuals with autism face. But it does allow for more precise, compassionate, and supportive approaches. We must separate the identity of being autistic from outdated ideas of dysfunction. The current generation of individuals with autism has the chance to grow up in a society better equipped to understand them.

After 50 years of shifting definitions and improved awareness, we are finally at a point where the question is no longer “What causes autism?” but “How can we best support people with autism in a diverse world?”