Dissecting the exposome linked to mental health outcomes can help identify potentially modifiable targets to improve mental well-being. However, the multiplicity of exposures and the complexity of mental health phenotypes pose a challenge that requires data-driven approaches.

Guided by our previous systematic approach, we conducted hypothesis-free exposome-wide analyses to identify factors associated with 7 psychiatric diagnostic domains and 19 symptom dimensions in 157,298 participants from the UK Biobank Mental Health Survey. After quality control, 294 environmental, lifestyle, behavioral, and economic variables were included. An Exposome-Wide Association Study was conducted per outcome in two equally split datasets. Variables associated with each outcome were then tested in a multivariable model.

Across all diagnostic domains and symptom dimensions, the top three exposures were childhood adversities and traumatic events. Cannabis use was associated with common psychiatric disorders (depressive, anxiety, psychotic, and bipolar manic disorders), with ORs ranging from 1.10 to 1.79 in the multivariable models. Additionally, differential associations were identified between specific outcomes—such as neurodevelopmental disorders, eating disorders, and self-harm behaviors—and exposures, including early life experiences (being adopted), lifestyle (time spent using computers), and dietary habits (vegetarian diet).

This comprehensive mapping of the exposome revealed that several factors, particularly in the domains of those previously well-studied were shared across mental health phenotypes, providing further support for transdiagnostic pathoetiology. Our findings also showed that distinct relations might exist. Continued exposome research through multimodal mechanistic studies guided by the transdiagnostic mental health framework is required to better inform public health policies.

Schizophrenia negatively affects quality of life (QoL). A handful of variables from small studies have been reported to influence QoL in patients with schizophrenia, but a study comprehensively dissecting the genetic and non-genetic contributing factors to QoL in these patients is currently lacking.

We adopted a hypothesis-generating approach to assess the phenotypic and genotypic determinants of QoL in schizophrenia.

The study population comprised 1119 patients with a psychotic disorder, 1979 relatives and 586 healthy controls. Using linear regression, we tested >100 independent demographic, cognitive and clinical phenotypes for their association with QoL in patients. We then performed genome-wide association analyses of QoL and examined the association between polygenic risk scores for schizophrenia, major depressive disorder and subjective well-being and QoL.

We found nine phenotypes to be significantly and independently associated with QoL in patients, the most significant ones being negative (β = −1.17; s.e. 0.05; P = 1 × 10–83; r2 = 38%), depressive (β = −1.07; s.e. 0.05; P = 2 × 10–79; r2 = 36%) and emotional distress (β = −0.09; s.e. 0.01; P = 4 × 10–59, r2 = 25%) symptoms. Schizophrenia and subjective well-being polygenic risk scores, using various P-value thresholds, were significantly and consistently associated with QoL (lowest association P-value = 6.8 × 10–6). Several sensitivity analyses confirmed the results.

Various clinical phenotypes of schizophrenia, as well as schizophrenia and subjective well-being polygenic risk scores, are associated with QoL in patients with schizophrenia and their relatives. These may be targeted by clinicians to more easily identify vulnerable patients with schizophrenia for further social and clinical interventions to improve their QoL.