Does emotional suffering fuel creativity—or does creative work increase mental-health problems? Popular stories often give a simple answer.
A study published in Molecular Psychiatry on 8 October 2026 analysed 14,658 Swedish twins and reveals why that question needs to be separated into several very different claims.
Quick answer
Does the Swedish twin study prove that creativity causes depression?
No. Artistic and scientific creativity showed different mental-health associations. Many links weakened within identical twin pairs, pointing toward shared familial influences. One possible path from earlier depressive symptoms to later openness does not prove that depression produces creative talent.
Key takeaways
- The cohort included 14,658 Swedish twin individuals; 6,419 completed both survey waves.
- Artistic and scientific creativity showed different associations with mental-health outcomes.
- Many associations weakened in identical-twin comparisons, consistent with shared family and genetic influences.
- A possible link from earlier depressive symptoms to later openness remained; openness is not artistic skill.
- The research does not show that depression creates genius or that creativity itself causes psychiatric illness.
The myth of the troubled creative genius
The image of the brilliant artist who must suffer for their work is familiar from biographies, films and popular psychology. It is also a claim about health that deserves careful testing. If some creative people report more depressive symptoms, does their creativity make them unwell? Could mental-health difficulties change the way they approach ideas and art? Or do the same inherited influences contribute to both?
A new Swedish twin study in Molecular Psychiatry examined these possibilities with unusually varied definitions of creativity. The central finding is not that creativity causes depression. It is that the answer depends heavily on what kind of creativity is being measured, what health outcome is considered and whether relatives with shared genetic backgrounds are compared.
Who took part in the Swedish research?
Penghao Xia and colleagues analysed 14,658 twin individuals from the STAGE cohort of the Swedish Twin Registry. Participants were born between 1959 and 1985 and completed surveys in 2012–2013 and 2022–2023; 6,419 people participated in both waves. Some participants also provided DNA samples for polygenic analyses.
The researchers linked participants to Swedish national patient records covering diagnoses through the beginning of 2017. The study thus combined self-reported personality, creative achievements and symptoms with clinical records, occupational information and genetic indicators. Not every measure was available for every person at both waves, which matters when interpreting any single result.
What did the researchers mean by creativity?
Creativity is not a single laboratory measurement. Artistic achievement covered music, writing, dance, theatre and visual arts. Scientific achievement covered science and invention. The study also distinguished artistically creative employment from other creative occupations, such as certain roles in science and engineering.
Another measure, divergent thinking, asked participants to generate ten words that differed as much as possible in meaning. Researchers also included openness to experience, a personality dimension associated with curiosity and receptiveness to new ideas. Openness may correlate with creative interests, but it is not interchangeable with artistic ability, professional achievement or originality.
What counted as a mental-health outcome?
The study included questionnaire-based depressive symptoms and psychosis proneness at the first survey wave. Those are not the same as a diagnosis. Separately, the team analysed registered major depressive disorder, anxiety disorder, stress-related disorders, schizophrenia and bipolar disorder.
The register outcomes reflect diagnoses recorded in specialist health care, not every person experiencing distress. Symptoms were not measured identically across both waves, and the researchers could not run every possible longitudinal model for every outcome. Headlines that collapse all these measures into 'depression' obscure important distinctions.
Artistic creativity and scientific creativity did not behave alike
In ordinary statistical analyses, greater artistic achievement and artistically creative employment were associated with more depressive symptoms and psychosis proneness in some comparisons. Greater openness was associated with psychosis proneness. Certain artistic indicators also predicted higher subsequent rates of some registered psychiatric diagnoses.
By contrast, scientific creativity was more often unrelated to those problems, while measures of occupational creativity and non-artistically creative employment sometimes showed associations in the opposite direction. This is not a ranking of artists against scientists, and it does not mean a particular profession protects mental health. The findings demonstrate why one broad label, 'creative people', can be misleading.
What happened when identical twins were compared?
The investigators used co-twin control analyses for associations that passed their initial statistical screening. By comparing identical twins who differed in a measured trait, they could account for inherited DNA and many shared family experiences that might influence both creativity and health.
Most associations became weaker and were no longer statistically significant in these within-pair models. This pattern is consistent with shared familial influences contributing to the apparent links. It does not prove that every effect is absent: twin pairs still differ in experiences, measurement can be imprecise, and some rare diagnoses provided too few discordant pairs for meaningful comparisons.
The finding that needs the most careful interpretation
One longitudinal result remained in the identical-twin comparison: twins with higher depressive-symptom scores at the first wave tended to report greater openness at the second wave than their co-twin. The authors interpreted this as a possible influence of earlier symptoms on later openness, while acknowledging the limits of observational evidence.
Crucially, openness is a personality dimension—not a measure of artistic mastery. The finding does not show that depression creates genius, that mental illness improves creativity or that anyone benefits from untreated symptoms. It also does not establish a clinical intervention or a guaranteed cause-and-effect pathway.
Where do genes enter the picture?
The team used polygenic scores, which summarize small statistical associations across many genetic variants, and genetic correlations from genome-wide research. These methods suggested overlap between some genetic influences associated with mental-health outcomes and those associated with creative traits.
A genetic correlation is not a gene for creativity, a diagnosis or a destiny. It cannot show whether one trait directly causes another. Polygenic scores are also sensitive to the populations and discovery studies used to construct them, and they cannot accurately predict an individual's creative career or psychiatric future.
What the study can and cannot establish
The analysis benefits from a large Swedish twin cohort, multiple measures, two survey waves, national registers and identical-twin comparisons. It also has limits: respondents were adults born between 1959 and 1985; some measures relied on self-report; openness questionnaires changed between waves; and the sample completing both surveys was smaller than the headline cohort.
The researchers controlled false discoveries across many statistical tests, but some models had limited statistical power. Rare diagnoses and creative occupations can be especially difficult to study within identical twin pairs. A non-significant result should not be translated into certainty that no association exists. Nor can observational results demonstrate that making art treats or causes mental illness.
What this means for everyday creativity
There is no reason to view depression as a necessary price of artistic expression. Creative activities can have personal, cultural and social value regardless of how they relate to a particular study's psychiatric measures. People who feel distressed deserve support without being told that symptoms are the source of their talent.
For readers interested in the science, the more useful question is not whether 'creatives' are inherently at risk, but which facets of creativity, personal history, environments and genetic influences shape observed associations. This study advances that question; it does not settle it for every person or occupation.
Source notes
The source has been verified and editorially reviewed for this article. The limitations below show which level of conclusion the sources support.
- [xia-creativity-mental-health-swedish-twins-2026] Unraveling the multifaceted relationship between creativity and mental health with longitudinal analyses of genotyped twins Penghao Xia; Laura W. Wesseldijk; Yi Lu; Sverker Sikström; Örjan de Manzano; Vivien Vach; Fredrik Ullén; Miriam A. Mosing. Molecular Psychiatry, 2026. Evidence type: Genetically informed longitudinal observational STAGE study of 14,658 Swedish twins with identical-twin within-pair comparisons, register-linked diagnoses and polygenic analyses Limitation: Published 8 October 2026. Artistic and scientific creativity showed different mental-health associations; many apparent links attenuated in monozygotic co-twin comparisons. The observed depressive-symptom-to-openness association remained, but does not establish a clinical causal effect. PubMed DOI
Editorial source review
This section shows how the article's key factual claims are linked to the source.
Phrasings that require caution
- Separate self-reported symptoms from registered psychiatric diagnoses.
- Do not equate openness with artistic talent or frame depression as a route to genius.
- Non-significant within-pair results do not prove zero effect.
- Polygenic associations are not individual diagnoses or reliable predictions.
- Do not claim TwinPare can diagnose depression or measure creative ability.
| ID | Claim | Source support | Caution |
|---|