Effects of lorazepam on deductive reasoning

Benzodiazepines slow reasoning performance, but it is still unknown which phase of reasoning is affected and whether this effect is present for different types of relations between entities in reasoning problems.

We investigated which phases of deductive reasoning are affected by lorazepam and whether this effect varies according to the type of relations in deductive reasoning problems.

This was a double-blind, crossover design study of acute oral doses of lorazepam (2 mg) and placebo, using young healthy volunteers. We focused on response delay of three separable phases of deductive reasoning and matched working memory tasks (that involved only maintenance of information) the premise processing phase, the premise integration phase, and the validation phase, in which reasoners decide whether a conclusion logically follows from the premises (reasoning task) or is identical to one of the premises (maintenance task). Type of relations in the premises was also manipulated. We employed material that was difficult to envisage visually and visuospatially ("subiconic") and material easy to envisage visually or visuospatially.

Lorazepam slowed response as memory load increased, irrespective of type of relations. It also specifically slowed validation in reasoning problems with visual relations, an effect that disappeared after subtraction of maintenance scores, and increased validation time in problems with subiconic relations, which remained after this subtraction.

Acute lorazepam administration affected reasoning in two ways: it slowed processing nonspecifically when working memory demands increased and augmented validation time depending on the difficulty in generating and/or manipulating mental representations by the central executive.