Send to

Choose Destination
Neuroimage. 2009 Apr 15;45(3):1020-32. doi: 10.1016/j.neuroimage.2008.12.034. Epub 2008 Dec 30.

Lack of generalizability of sex differences in the fMRI BOLD activity associated with language processing in adults.

Author information

Department of Neurology, Washington University School of Medicine, St. Louis, MO 63110, USA.


A lack of consensus exists as to whether there are sex differences in the fMRI BOLD signal correlates of language processing in the human brain. Here, whole-brain fMRI was used to examine the neural activity of 46 adults performing one of two sets of language tasks. Conservative quantitative and qualitative criteria identified a handful of statistically significant regions of "sex difference" within each task separately. When each of the two sets of regions was investigated in the group of subjects performing the other task set, however, most of the identified "sex differences" failed to generalize. Identical analyses of the same subjects divided into sex-matched pseudorandom control groups for each task set separately revealed that it is possible to observe a similar number of statistically significant regions of "group difference" in the task-associated BOLD signal, even when the groups do not differ on any of the measured behavioral parameters, or any obvious demographic characteristic. Together, these results suggest that one should be cautious when interpreting studies that purport to have identified regions of difference between groups, whether those groups are divided by sex or by any other criterion. In particular, generalization or replication of a result in independent data sets is necessary for establishing conclusive support for any hypothesis about differences in brain function between groups.

[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Elsevier Science Icon for PubMed Central
Loading ...
Support Center