Cannabis cues grip the brain harder than users feel they do

Craving is one of the engines of continued drug use, and the standard way to study it is cue-reactivity: show someone reminders of the drug and measure how their mind and body respond. For alcohol and nicotine this approach is well established, but for cannabis, one of the most widely used drugs in the world, the overall size of the effect had never been pinned down. Knowing whether cannabis cues genuinely provoke measurable responses, and which measures pick them up, matters for understanding why habits persist and why some people find moderation harder than they expect.

The researchers swept five databases for studies published between January 2000 and December 2014 in which regular cannabis users were exposed to cannabis cues and their responses compared against neutral cues or a baseline assessment. Fourteen studies met the criteria, but data could not be obtained for two, leaving 12 studies with 393 participants between them, ranging from 12 to 97 people per study, with mean ages spanning 16.4 to 32.6 years. Responses included heart rate, skin conductance, skin temperature, blood pressure, a brain-wave measure called the Pz P300, and self-reported craving on standard questionnaires, all combined in a single statistical model.

On average, users showed moderate cue-reactivity, whether their cue responses were compared with a baseline period or with neutral cues, and the main model showed no evidence of publication bias. The picture varied sharply by measure: the brain-wave (Pz P300) response was large, while heart rate showed a trivial change that was not statistically reliable. Those subgroup results came with a warning, though, because the individual studies disagreed with each other considerably on the physical measures. The authors' overall reading was that users' attentional pull towards cannabis cues runs much stronger than the craving they report feeling.

This was, at the time, the first attempt to put a single number on cannabis cue-reactivity by pooling all the eligible studies, and it establishes that the phenomenon is real and moderate in size rather than marginal. The gap it exposes is intriguing: the brain's attention-related response dwarfed people's subjective sense of wanting the drug, suggesting self-report alone understates how reactive regular users are. What the analysis cannot say is whether stronger cue-reactivity predicts heavier use, relapse after quitting, or treatment outcomes, and the disagreement between studies on physical measures means the exact sizes remain uncertain.

If you use cannabis regularly, this is a reason to think that the sights, smells and situations linked to the drug are doing more work on you than your conscious craving suggests. Someone who feels only mild desire may still show a strong automatic pull towards cannabis-related cues, which could help explain why cutting back can feel harder in familiar settings than expected. It is informational rather than a strategy: the study measured reactions, not whether avoiding cues changes use. But it reframes craving as something your attention can register before you consciously feel it.