Alexithymia, Autism, and Why 50% Is Not 100%

In discussions about autism and emotional life, one statistic is frequently repeated: around 50% of autistic people meet criteria for alexithymia, compared with roughly 10% of neurotypical individuals. It is a valid statistic, grounded in research, and worth taking seriously. But it is often used in a way that quietly distorts what it actually tells us.

Because 50% is not 100%. And that difference matters.

The moment this distinction is ignored, statistics begin to slide into assumptions. What was meant to describe a partial overlap starts to behave like a defining trait. Alexithymia becomes confused with autism itself, and difficulty identifying emotions is mistaken for an absence of emotional life. From there, behaviours are interpreted as evidence, rather than as signals that require understanding.

This is where the real damage happens.

Alexithymia does not describe whether emotions exist. It describes access, the ability to recognise, differentiate, and translate internal emotional states into conscious awareness and language. A person may feel intensely and yet struggle to name what is happening inside them. The body registers emotion; the mind lacks a map.

When this distinction is missed, emotional experience is inferred from behaviour. Silence is read as emptiness. Directness becomes coldness. Regulation differences are interpreted as lack of empathy. The problem is not the data; it is the leap from population-level statistics to individual conclusions.

The fact that roughly half of autistic people do not meet criteria for alexithymia should already be enough to stop generalisation. Yet even within the group that does, alexithymia is not fixed, not universal, and not identical in expression. It is shaped by context, safety, developmental history, trauma, environment, and relational experience. Treating it as a static trait flattens complexity into stereotype.

This is why checklist-based assessments fail so often. They reward pattern recognition over understanding and replace curiosity with certainty. Two people can show the same external behaviour and live entirely different internal realities. Without individual evaluation, those differences are erased.

The ethical importance of the “not 100%” cannot be overstated. It forces a shift in posture: from categorising to listening, from labelling to inquiry, from explaining behaviour to understanding experience. It reminds us that diagnosis should never substitute relationship, and statistics should never replace attention.

What this data should reinforce is not caution about emotions, but responsibility in how we assess them. Every autistic person deserves to be approached as an individual, not as a probability. Emotional life cannot be inferred from surface behaviour, and emotional access can change when there is time, trust, and safety to develop a shared language.

When we forget this, we stop seeing people and start seeing averages. And averages, however useful in research, are incapable of holding a human life.

Fifty percent is not one hundred percent. And respecting that difference is not just scientific accuracy, it is a matter of dignity. Ligia Koijen founder