Understanding the Limitations of the Odds Ratio in a Case-Control Study of the Association Between Breast Cancer and Exposure to Antipsychotic Drugs

Abstract

Evidence suggests that women with schizophrenia are less likely to be screened for breast cancer, more likely to suffer from breast cancer, and more likely to die of breast cancer than women without schizophrenia or general population controls. Antipsychotic drugs, and especially prolactin-raising antipsychotic drugs, have been suggested to increase the breast cancer risk, but the evidence has so far been inconclusive. Against this background, a recent, large, nationwide, case-control study in Finland examined the odds of previous prolonged exposure to prolactin-raising and prolactin-sparing antipsychotic drugs in women with schizophrenia who were (cases) versus were not (controls) diagnosed with breast cancer. The study found that, relative to < 1 year of antipsychotic exposure, breast cancer was associated with significantly increased odds of previous, prolonged (> 5 years) exposure to prolactin-raising antipsychotics. The associations were not statistically significant for prolactin-sparing antipsychotics. The study is critically examined from the perspective of interpretation of the odds ratio and its limitations in order to help readers understand how to better evaluate and generalize findings in case-control studies. This is necessary because results in case-control studies are often incorrectly interpreted, and the limitations of the odds ratios derived in such studies are often not recognized. It is concluded that the design and findings of the reviewed study could not allow readers to judge whether or not prolactin-sparing antipsychotics are associated with lower breast cancer risk than prolactin-raising antipsychotics. In contexts other than breast cancer risk, adverse consequences associated with prolactin elevation are well known, and avoidance or management of hyperprolactinemia is therefore desirable.