Sean X. Luo M.D., Ph.D.

Hooked on Patterns

This Is NOT Your Brain on Drugs

Popular coverage of neuroimaging results of addiction needs more care

Posted Nov 02, 2014

NYtimes: This is your brain on drugs.  It is an article that covers a new study coming out of Massachusetts General Hospital, showing a composite scan of brains of 20 pot smoking young adults, with a little red circle indicating a part of the brain that’s “most different” between the pot smokers and normal controls.

The article subsequently starts to string things together that can be highly misleading.  “Moderate marijuana use by healthy adults seems to pose little risk.”  Is that really true?  What exactly is moderate?  Is that true in every individual?   First it cites various sources in saying that THC content on average has increased, and then says, “higher potency may also accelerate addiction.”  But this statement is only backed by causal remarks.

Finally, it winds itself back to the original neuroimaging study, as reports “all smokers showed abnormalities in the shape, density and volume of the nucleus accumbens.”

Being pretty familiar with neuroimaging results in biological psychiatry, I was pretty damn 100% sure that that interpretation is patently false.  And I went back to the original article, and lo-and-behold it was patently false.

First of all, there is NO clinically meaningful normal “shape, density and volume of nucleaus accumbens.”  Radiologists do not measure these parameters in making diagnoses.  There’s no quantification of what is normal and what isn’t.  The paper simply shows that there is a particular value from the brains of pot smokers that correlates to how much pot you smoke.

Figure 1 adapted from [1]

Second of all, it’s false that “all smokers” showed “abnormalities.” (Figure 1).  It’s a pretty easy to read figure, and you don’t need a PhD to make an inference.  X-axis how much pot you smoke, Y-axis is a value (GMd) you extract from the brain.  The more pot you smoke, the higher that value you get.  The authors want us to interpret that value as some kind of abnormality.  But unless you are blind, you’d see that there is a high degree of overlap bwetween the pot smoking group(s) and the controls.  The effect size is not particularly impressive despite the small p values.

Could it be that some pot smokers have brains that look exactly the same as some normal controls?  The data seem to say so. 

Third of all, correlation does not imply causation.  But this point is always lost in popular coverage.  Is the pot smoking that causes the supposed brain change or is it the brain change that predispose people a risk to smoke more pot?

Finally it cites a controversial but well cited study from the Duke group on persistent cannabis use dropping 8 points on IQ in a large longitudinal study [2].  But the newspaper article does not cover the later analysis that shows that one explanation is a confounding factor of socioeconomic status: let’s say for some other reason individuals who smoke pot also are more liable to drop in their socioeconomic status.  It turns out the drop of socioeconomic status can also drop your IQ later in life: Rogberg’s analysis of a similar cohort apparently did not replicate the original results [3].   Causal interpretation is flawed and immature.  Of course, there are a bunch of replies, and Terrie Moffitt wrote a reply [4] that seems to suggest at least in the Duke cohort, the confounding effect cannot be the sole explanation. So the jury is really still out on this issue, but it’s fair to say that there is a risk of IQ drop in some individuals.  We don’t know who they are.  We don’t know how to find them.  NYtime’s blanket statement is disturbingly inaccurate. 

Just a very quick exegesis of an article shows that the conclusions drawn are somewhere between inaccurate to patently false.  While I have no objections per se to the scientific results of any of the papers cited, all of which I think are important and interesting, the popular coverage interpretation are simplistic, politically driven, and sometimes just false.  The relationship between marijuana, psychosis, cognitive function and brain structure is highly complex and probably differ to a substantial degree amongst different individuals.  These popular articles reinforce the notion that you are either an addict with a different brain or a normal person with a normal brain.  I expect more from the New York Times.  This kind of journalistic practice creates unnecessary stigma and really needs to stop.

[1] Gilman JM et al., 2014 Cannabis Use Is Quantitatively Associated with Nucleus Accumbens and Amygdala Abnormalities in Young Adult Recreational Users, J Neurosci., 34: 5529 

[2] Meier et. al., 2012, Persistent cannabis users show neuropsychological decline from childhood to midlife, PNAS, 109:2657

[3] Rogeberg, O 2013, Correlations between cannabis use and IQ change in the Dunedin cohort are consistent with confounding from socioeconomic status PNAS, 110: 4251

[4] Moffitt et al., PNAS 110:980

About the Author

Sean X. Luo, M.D., Ph.D., is a physician-scientist working at Columbia University and The New York State Psychiatric Institute.

More Posts