The idea that creativity comes with a psychological price survives because it turns complicated lives into a memorable bargain: exceptional art in exchange for suffering. A new study in *Molecular Psychiatry*, published October 8, asks a harder question. When creative people show more mental-health problems, is creativity itself involved—or do both reflect influences that run through families?
The researchers analyzed 14,658 Swedish twins, drawing on two survey waves roughly a decade apart, psychiatric diagnoses from national registers, and genetic data from a subsample. Their result is not a neat vindication or demolition of the “mad genius” story. It is more useful than that: the apparent relationship changes according to what counts as creativity, what counts as mental ill health, and whether twins are compared with strangers or with each other.
Creativity stopped behaving like a single trait
Many studies treat creativity as though painting, scientific problem-solving, personality and job title were interchangeable. This team separated artistic and scientific achievement, artistically and non-artistically creative employment, occupational creativity, openness to experience and divergent thinking. Those measures overlapped only modestly in some cases, even though they were reasonably stable over time.
That distinction changed the pattern. In population-level analyses, higher artistic achievement and artistically creative employment were associated with more depressive symptoms and psychosis-like experiences. Higher artistic achievement also predicted a greater subsequent risk of schizophrenia, anxiety and stress-related diagnoses. Openness—a broad personality trait that includes imagination, curiosity and aesthetic sensitivity—showed several positive associations with mental-health problems.
Scientific creativity told a different story. Scientific achievement, broadly creative occupations and non-artistic creative employment tended to show no association, or sometimes an association with fewer mental-health problems. This does not mean science protects mental health or art damages it. It means the label “creative” was hiding distinct domains whose statistical relationships pointed in different directions.
That helps explain why earlier research has looked contradictory. A 2017 systematic review and meta-analysis found that results varied depending on whether researchers compared clinical and nonclinical groups, creative and noncreative groups, or continuous symptom and creativity scores. The question and measurement strategy can shape the answer.
The twin comparison weakened the headline
Ordinary observational comparisons cannot easily separate an individual exposure from family background. Twins offer a stricter test. If one twin is more artistically accomplished than the other, does that twin also show poorer mental health? Such co-twin analyses partly control for shared upbringing and genes; identical twins provide the strongest version because they share essentially all their inherited DNA.
Most of the study’s associations became smaller and statistically non-significant in these within-family comparisons. The genetic analyses pointed in the same general direction: some genetic influences appeared to contribute to both artistic creativity and mental-health traits. In other words, shared familial or genetic factors may help produce the population-level link. The finding challenges the simple causal claim that creativity creates illness, or that illness creates creativity.
One result resisted that general pattern. Twins reporting more depressive symptoms than their co-twin at the first wave later scored higher on openness, even after the analysis accounted for earlier openness. The authors describe this as a possible causal relationship. “Possible” matters. Openness is not creative achievement, depressive symptoms are not identical to a diagnosis of depression, and a co-twin design cannot eliminate every nonshared influence. The study also found that a previous major-depression diagnosis did not predict higher openness once family factors were controlled.
Other limits narrow the conclusions. Mental-health registry data ended in early 2017, while the second survey occurred in 2022–23. Some measures relied on self-report, rare diagnoses left too few discordant twin pairs for certain analyses, and Swedish twins born between 1959 and 1985 may not represent other generations or societies. Even a non-significant co-twin result cannot prove there is no causal effect; it may reflect limited statistical power.
Making art is not the same question as being an artist
The findings might seem to clash with evidence that participating in the arts can support well-being. They do not. The World Health Organization’s evidence review on arts and health examined how artistic activities may prevent illness, promote health or complement care. This twin study asked whether stable traits, achievements and occupations co-occur with symptoms or diagnoses. An activity can be helpful without every highly artistic person enjoying better mental health; those are different comparisons.
The practical lesson is therefore neither to romanticize distress nor to distrust creativity. Mental illness is not a credential for original work, and creative achievement is not evidence of hidden pathology. The strongest conclusion from the new research is that there is no single creativity–mental health link. Artistic and scientific domains differ, family factors matter, and one dramatic cultural story cannot carry all that complexity. The “mad genius” myth does not survive the twin comparison intact—and psychology is better for losing the slogan.
