This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
—To address the first question, we did have a lower CI on our male subjects that was equal to 1.00. Our P value was equal to.05. There is still a one in 20 chance that the data may be wrong. We chose to report our result as statistically significant: perhaps reporting the P value in this case would have been less confusing.We analyzed the data as a retrospective cohort. It is for this reason that we used RR instead of OR, as would be more commonly used in a case-control study.Again, the study was ecological in nature. The RRs were small and we did not get individual data as other risk factors for hip fracture. The results are interesting, but certainly not conclusive.
Christa Danielson, Joseph M. Lyon. Hip Fractures and Fluoridation-Reply. JAMA. 1993;269(16):2087. doi:10.1001/jama.1993.03500160053025