A study published in the American Journal of Epidemiology sought to determine if the risk of breast cancer is elevated for night-shift workers. Reuters reports, [Wong-Ho Chow, a researcher with the National Cancer Institute] followed more than 70,000 women in Shanghai, inquiring about their work and cancer status. Earlier research in animals had suggested that disturbances in the light-dark cycle could knock certain chemicals in the body -- for instance, melatonin and estrogen -- off balance, and as a result increase cancer risk. ¦ But the risk didn't increase for women who worked nights, regardless of how long they had done so.
They had this hypothesis that working night shifts increases cancer risk based on animal studies, says Stier. So they decided to test the hypothesis on humans, and it turns out that there is no connection in this case. This is a good example of the fact that you can t necessarily build a hypothesis for humans based on animal data. We re not saying you shouldn t do animal tests, just that you have to do human tests before you can know if the study is relevant to human health.