The nine Bradford Hill criteria are used to further assess the possibility that selenium deficiency may be a risk factor in si.hiwvhrenia. This hypothesis appears compatible with the known link between industrialization and schizophrenia since air pollution increases soil acidity, reducing selenium's bioavailabiliiy. Several feasible biological mechanisms far a selenium deficiencyóschizophrenia association are also identified, including prostaglandin imbalances, viral mutation and excess 12-HPETE production.

Evidence of a strong negative correlation between activity of the selenoertsyme glutathwmperoxidase and schizophrenic brain atrophy is also revieuied. Nine surveys, conducted between 1880 and 1963 indicate a continuing negative relationship between selenium levels in fodder crops and United States schizophrenia prevaleiice. Such variations in the prevalence of schizophrenia, in state and'county mentril hospitals suggest a selenium relatedrelative risk of 1.77:1. International similarities in the spatial distributions of'schizophrenia and celiac disease, cancer of the esophagus and multiple sclerosis (all of which appear to involve selenium inadequacy as either a cause or an effect) further support a role for selenium deficiency in schizophrenia. It is concluded that the available evidence warrants the careful testing of selenium supplementation in dye prevention and possible treatment of schizophrenia.