Statistical issues in the analysis of low-dose endocrine disruptor data

The National Institute of Environmental Health Sciences (NIEHS) and the U,S , Environmental Protection Agency (U.S. EPA) recently cosponsored the Endoc rine Disrupters Low-Dose Peer Review. The purpose of this meeting was to ex amine data supporting the presence or absence of low-dose effects of endocr ine disrupters in specific studies and then to evaluate the likelihood and significance of these and/or other potential low-dose effects for humans. A ll invited speakers agreed to provide their raw data in advance of the meet ing to a Statistics Subpanel, which was asked to reevaluate the authors' ex perimental design, data analysis, and interpretation of experimental result s, The purpose of this statistical reevaluation was to provide an independe nt assessment of the experimental design and data analysis used in each of the studies and to identify key statistical issues relevant to the evaluati on and interpretation of the data. This paper presents a summary of the Sta tistics Subpanel's evaluation. Specific examples are presented to illustrat e problems that arose in the experimental design and data analysis of certa in studies, The statistical principles and issues that are discussed in thi s paper are not unique to endocrine disrupter studies and should provide im portant guidelines regarding appropriate experimental design and statistica l analysis for other types of laboratory investigations.