Effects of alcohol and hypnosedative drugs on digit-symbol substitution: comparison of two different computerized tests

Computerized symbol-digit substitution tests (SDSTs) and 'yes/no' digit-symbol substitution tests (YNDSTs), together with a digit-digit copying test (DDCT) were compared in a double-blind crossover study for hypnosedative and alcohol effects in 12 young healthy subjects. The tests were carried out before and 1, 3.5 and 5 h after intake of 15 mg diazepam, 30 mg oxazepam, 7.5 mg zopiclone,15 mg zolpidem or 0.65+0.35 g/kg ethanol, these impaired performance by varying extents. The drug-induced decrements were similar in the YNDSTs and SDSTs; the errors were subject-related and increased with alcohol intake in the YNDSTs but not in the SDSTs. Pooled baseline values in YNDSTs correlated well (Pearson) with those in the SDSTs. The SDSTs also correlated well with the DDCTs.The same correlations were seen when comparing absolute performances after ethanol and hypnosedatives. When using the Δ-values (changes from baseline), the correlation betweenYNDSTs and SDSTs was high after placebo but low after alcohol and hypnosedatives. When using ANCOVA to measure the responses to drugs (Δ-drug - Δ-placebo), the correlation between YNDSTs and SDSTs was fairly significant after most but not all active drugs. Subjective visual analogue variables did not correlate with objective performances. By comparing absolute performances with hypnosedatives we conclude that YNDSTs and SDSTs are similar, with the Δ-values indicating differences between them.