A white-matter measure correlated with Stroop interference among abstinent methamphetamine users even though the corrected imaging difference between users and controls was not statistically significant. Those two findings answer different questions and should not be merged into a claim of proven brain damage.
Salo and colleagues compared 37 currently abstinent adults with a history of methamphetamine abuse with 17 non-substance-abusing controls. Participants completed a computerized Stroop task and diffusion tensor imaging. The imaging assessed water-diffusion characteristics in the genu and splenium of the corpus callosum. Fractional anisotropy, or FA, is an imaging index, not a direct count of damaged nerve fibers.
The methamphetamine group showed greater reaction-time interference on the Stroop task (p = 0.04). Within that group, lower genu FA was associated with greater interference after Bonferroni correction (p = 0.04). However, the between-group genu FA comparison had a corrected p value of 0.09. Calling that a statistically significant group difference would overstate the result.
Splenium FA did not significantly differ between groups, and several other diffusion indices also showed no significant group differences. Although the genu correlation was significant and the splenium correlation was not, the direct test comparing their slopes was nonsignificant (p = 0.155). A significant result in one region and a nonsignificant result in another do not automatically prove that the regions differ.
This was an observational human study, not an experiment assigning methamphetamine exposure. Pre-existing differences, prior substance use and smoking were potential influences. The Stroop comparison also weakened when reading-based estimated ability was included as a covariate. Neither DTI nor the group association measured current drug concentration or established impairment during a specific event. The useful conclusion is narrower: one anterior callosal imaging measure tracked one aspect of cognitive control within this sample.
For a separate question, see sensory brain responses during cognitive-control testing.
Read the original study (DOI: 10.1016/j.biopsych.2008.08.004).