A correlation between a diffusion MRI measure and task performance does not establish that a scan diagnoses methamphetamine-caused injury. A 2009 study illustrates this distinction: its principal brain–behavior association reached significance, while the corrected group difference in the relevant imaging measure did not.
What was measured?
Salo and colleagues compared 37 participants with methamphetamine dependence and 17 controls. The methamphetamine group had been abstinent for at least three weeks. Participants completed a computerized Stroop task, and diffusion tensor imaging assessed the anterior genu and posterior splenium of the corpus callosum.
Fractional anisotropy, apparent diffusion coefficient and diffusion eigenvalues describe aspects of water diffusion in tissue. They are not direct counts of damaged axons or a measurement of methamphetamine in blood. No dose was administered in this observational comparison.
The behavioral and imaging comparisons differed
The methamphetamine group showed greater Stroop reaction-time interference, with p = 0.04 and a reported standardized group effect of 0.66. The result persisted with education as a covariate but became a nonsignificant trend when a reading-based estimate of prior ability was included.
The genu fractional-anisotropy group comparison was nonsignificant after correction, p = 0.09. The splenium comparison was also nonsignificant, p = 0.88. Group comparisons involving apparent diffusion coefficient and the examined eigenvalues were not significant in either callosal region.
A significant association is a different result
Within the methamphetamine group, greater Stroop interference was associated with lower genu fractional anisotropy after correction. The corresponding splenium association was nonsignificant. However, a direct comparison between the regional slopes was not significant.
Thus, “significant in the genu, nonsignificant in the splenium” is not equivalent to a demonstrated difference between the two associations. Nor does a within-group correlation turn the nonsignificant between-group imaging comparison into a positive one.
The study did not find significant relationships between the imaging measures and years of methamphetamine use or months of abstinence. These null results limit simple exposure-duration or recovery interpretations.
What cannot be inferred?
The groups differed in education and reading-based ability estimates. Pre-existing differences, other exposure histories and selection into treatment remain possible influences. The cross-sectional design did not establish whether methamphetamine caused an individual’s imaging pattern.
The paper supplies historical evidence about a particular brain–behavior relationship. It does not validate an impairment cutoff, diagnose permanent injury, determine acute intoxication or establish a person’s driving ability. An individual scan requires an interpretation grounded in its actual clinical and technical context.
Primary scientific source
Salo R, Nordahl TE, Buonocore MH, et al. Cognitive Control and White Matter Callosal Microstructure in Methamphetamine Dependent Subjects: A DTI Study. Biological Psychiatry. 2009;65(2):122–128. doi:10.1016/j.biopsych.2008.08.004. Historical human observational research.
Leave a Reply