Gene

Quantum mechanics

Biology

Causal inference

Trait

Psychiatry

Mendelian randomization

Computer science

Genetic correlation

Single-nucleotide polymorphism

Genetics

Physics

Correlation

Bioinformatics

Schizophrenia (object-oriented programming)

Psychology

Mathematics

Genetic association

Genotype

Internal medicine

Geometry

Genetic variants

Medicine

Programming language

Observational study

Clinical psychology

Genetic variation

Causality (physics)

Pathology

Science Advances

There is a long-standing interest in exploring the relationship between blood-based biomarkers and psychiatric disorders, despite their causal role being difficult to resolve in observational studies. In this study, we leverage genome-wide association study data for a large panel of heritable serum biochemical traits to refine our understanding of causal effect in biochemical-psychiatric trait pairings. We observed widespread positive and negative genetic correlation between psychiatric disorders and biochemical traits. Causal inference was then implemented to distinguish causation from correlation, with strong evidence that C-reactive protein (CRP) exerts a causal effect on psychiatric disorders. Notably, CRP demonstrated both protective and risk-increasing effects on different disorders. Multivariable models that conditioned CRP effects on interleukin-6 signaling and body mass index supported that the CRP-schizophrenia relationship was not driven by these factors. Collectively, these data suggest that there are shared pathways that influence both biochemical traits and psychiatric illness.

Genetic estimates of correlation and causality between blood-based biomarkers and psychiatric disorders