A pilot experiment found no overall amphetamine benefit on spatial or digit-span memory, while psychosis-like experience ratings increased. A correlation between those changes did not prove a clinical psychotic disorder or its causal pathway.
Kassim and colleagues conducted two separate randomized double-blind placebo-controlled crossover pilots. Twenty participants received d-amphetamine 0.45 mg/kg in one experiment; another 20 received caffeine 200 mg. Forward spatial and digit-span performance was assessed at several delays, and ratings sampled psychosis-like experiences. The drugs were not compared within the same participant cohort.
Neither experiment demonstrated an overall drug improvement in spatial or verbal span. Amphetamine increased a composite psychosis-like experience score, and greater increases were associated with worse spatial-memory change, particularly at the longest delay. That relationship was an association between measured changes. It does not demonstrate that the rated experiences mediated a drug effect on memory, especially when the overall memory comparison was nonsignificant.
Both pilots were small and used forward-recall tasks rather than every type of working-memory demand. The between-cohort drug comparison also limits ranking amphetamine against caffeine. Psychosis-like rating changes in screened volunteers are not equivalent to diagnosing psychosis in an individual, and cannot be transferred automatically to chronic methamphetamine use. A null average effect likewise does not establish that every person’s memory was unchanged. The defensible account retains the negative task result, the subjective-rating effect and the uncertainty about their relationship as separate findings.
See why repeated task performance need not show a treatment-specific executive benefit.