Nikolaos A. Patsopoulos, M.D., of the University of Ioannina School of Medicine, Ioannina, Greece and colleagues evaluated a large sample of prominently claimed sex differences for genetic effects and whether these claims were methodologically strong or were made based on selected and/or suboptimal analyses and with insufficient or questionable documentation. From a database search the authors identified 77 articles with 432 sex-difference claims.
Of these claims, 286 (66.2 percent) sex comparisons were reported as being decided a priori (in advance of the study) and 68 (15.7 percent) were acknowledged to be post hoc (after the study) analyses; in the other 78 (18.1 percent), the analysis plan was unclear.
Appropriate documentation of gene-sex interaction was recorded in 55 claims (12.7 percent); documentation was insufficient for 303 claims and spurious (not valid) for the other 74. Data for reanalysis of claims were available for 188 comparisons. Of these, 83 (44.1 percent) were nominally statistically significant, and more than half of them (n = 44) failed to reach nominal statistical significance of a certain level. Of 60 claims with seemingly the best internal validity, only one was consistently replicated in at least two other studies. "The majority of these claims were insufficiently documented or spurious, and reporting of statistical interaction tests was rare" the authors write.
"We hope that our empirical evaluation will help sensitize clinicians, geneticists, epidemiologists, and statisticians who are pursuing subgroup analyses by sex or other subgroups on genetic associations. The pursuit of gene-sex interactions should not be necessarily abandoned. Ideally, sex differences should be based on a priori, clearly defined, and adequately powered subgroups.
Post hoc, discovery-based analyses are also of interest, but their post hoc character should be clearly stated in the manuscript. Both a priori and post hoc claims should be documented by interaction tests and proper consideration of the multiplicity of comparisons involved. Even then, results should be explained with caution and should be replicated by several other studies before being accepted as likely modifications of genetic or other risks."