Aug. 21, 2007 -- Many studies claim that a genetic mutation causes more disease in one sex than in another -- but most are wrong.
This provocative statement comes from Nikolaos A. Patsopoulos, MD, of the University of Ioannina School of Medicine, Greece, and colleagues. It's a shot across the bow for researchers tempted to make claims their studies do not actually support.
Patsopoulos and colleagues looked at 77 studies published in the medical literature. All of the studies prominently claimed to have discovered that a genetic variation is more likely to cause disease in one sex than in another.
The 77 studies made 432 claims of sex differences. The Patsopoulos team evaluated the statistical methods used in the studies and reanalyzed the data using appropriate statistical techniques.
The researchers were able to reanalyze data in 188 claims. In the end, fewer than half the claims (44%) had even borderline statistical significance -- that is, five-in-100 odds that the finding was chance. Only 44 of the 432 claims had "modest" statistical significance -- one-in-100 to five-in-100 odds the finding was chance.
Of the 60 most valid claims, only one was consistently validated in at least two other studies.
"At a minimum, the studies that we evaluated are probably among the ones in which authors were most certain about some, if not all, of the sex claims that they presented in their results," Patsopoulos and colleagues note. "Otherwise, they would not have drawn attention to the claims in the titles of their articles."
Analyzing research findings by sex is only the most common way researchers break down their findings. After data is collected, researchers often go back and analyze according to age, race, diet, lifestyle, and many other patient characteristics.
But these analyses are valid only when researchers planned for them in advance by including enough people in each subgroup to give the study the necessary statistical power.
"The vast majority of claimed subgroup differences are likely to be chance findings," Patsopoulos and colleagues conclude.
Their findings appear in the Aug. 22/29 issue of The Journal of the American Medical Association.