Two groups can make similar average choices on a laboratory task while differing in earnings or task-related brain activity. A 2014 methamphetamine study illustrates why these outcomes must be reported separately.
Task imaging and resting imaging had different denominators
Kohno and colleagues recruited 26 methamphetamine-dependent adults and 27 controls. One methamphetamine participant was excluded for excessive head motion, leaving 25 and 27 for the main imaging sample. Only 15 methamphetamine participants and 18 controls underwent the resting-state scan: 33 people, rather than all 52.
Participants were not seeking treatment. The methamphetamine group was studied after several days of abstinence, with urine testing used to verify abstinence from the specified substances. This was a case-control imaging study, not an administered-dose experiment or a test of acute intoxication.
How did the balloon task work?
Participants chose whether to pump a virtual balloon, increasing potential earnings by $0.25 per pump, or cash out. An explosion lost that balloon’s accumulated earnings. Colored balloons had different explosion contingencies, and control balloons provided a comparison condition.
Pump number indexed both increasing risk and increasing potential reward. The task therefore did not isolate one psychological process perfectly: evaluation, reward anticipation and choice could contribute to the measured signal.
What did the participants actually do?
The investigators found no significant group effect on pumping and no significant difference in average pumps before cashing out. Mean cash-out pump counts were 2.74 in the methamphetamine group and 2.84 in controls; the reported comparison was p = 0.180.
Controls nevertheless earned more overall: approximately $33.33 versus $30.15, with a significant comparison. It would be inaccurate to summarize this as uniformly greater pumping or greater willingness to take risk in the methamphetamine group.
Brain-response slopes differed
During pumping, right dorsolateral prefrontal activation increased more strongly with pump number in controls. Ventral-striatal modulation was greater in the methamphetamine group in the region-of-interest analysis. The prefrontal difference also appeared in the whole-brain analysis; no other significant whole-brain group differences or mean-activation differences were found in that comparison.
A steeper activation slope is distinct from a higher average signal. It describes how activity changes as the task changes, not a direct measurement of decision quality or brain injury.
Connectivity does not establish direction of control
The smaller resting-state sample showed greater midbrain connectivity with several regions in the methamphetamine group. Right dorsolateral prefrontal resting connectivity did not show a group difference.
Resting functional connectivity measures correlated signal variation. The authors expressly cautioned that it does not establish direction: a correlation between regions cannot by itself demonstrate that one region failed to control another.
Other limitations include small samples, differences in marijuana histories, overlapping control participants with earlier research, and the risk of inferring a mental process solely from an activated region. The findings are group-level associations, not an individual test of dangerousness, driving impairment or causal drug injury.
Historical primary source
Kohno M, Morales AM, Ghahremani DG, Hellemann G, London ED. Risky Decision-making: Prefrontal Function and Mesocorticolimbic Resting-state Connectivity in Methamphetamine Users. JAMA Psychiatry. 2014;71(7):812–820. doi:10.1001/jamapsychiatry.2014.399.
Leave a Reply