Background Previous studies have suggested that socio-environmental factors interact with genetic risk in the genesis of schizophrenia. This study aimed to investigate the relationship between income, income inequality and schizophrenia and its heterogeneity among different geographic scales and subgroups.

Methods We used data from the Second China National Sample Survey on Disability (2006). The sample consisted of 1 909 205 participants aged 18 years or older. Individuals who were suspected to be psychiatrically disabled were administered the WHO Disability Assessment Schedule, Version II and the International Statistical Classification of Diseases, Tenth Revision Symptom Checklist for Mental Disorders by trained clinical psychiatrists to diagnose schizophrenia. Gross household income per capita was used to calculate the Gini coefficient to measure income inequality. Multilevel logistic regression with cross-level interaction was applied to examine the association of income, income inequality and schizophrenia.

Results A total of 7 628 persons (0.40%) were identified as having schizophrenia. Income was independently associated with schizophrenia. At the province level, greater income inequality was significantly associated with a higher risk of schizophrenia (OR, 1.03; 95% CI 1.00 to 1.06), and no significant association was observed at the county level. The analysis with cross-level interaction showed that the association at the province level was most pronounced in the highest income quartile (OR, 1.02; 95% CI 1.00 to 1.03).

Conclusion The significant association between income and schizophrenia was consistent with the absolute income hypothesis. The adverse effect of income inequality on the risk of schizophrenia starts to operate at a larger area level, and it is more pronounced for the affluent population in China. This finding further supports the relative income hypothesis and social causation pathway for schizophrenia and calls attention to the vulnerability of high-income groups.