GLUCOCORTICOID-INDUCED IMPAIRMENT IN DECLARATIVE MEMORY PERFORMANCE IN ADULT HUMANS

Glucocorticoids (GCs) have a variety of effects on the brain including site-preferential, inhibitory effects on hippocampal neurons. In the case of dexamethasone (DEX), extended rather than single-dose treatmen t in vivo may be required for binding to brain rather than peripheral (e.g., pituitary) GC receptors and for maximizing other biologic effec ts in hippocampus (e.g., GC receptor downregulation, inhibition of glu cose transport). Based on the contributory role of hippocampal neurons in declarative memory performance, we investigated the cognitive cons equences of DEX treatment in normal adult human subjects, hypothesizin g a decrease in declarative memory performance after extended but not overnight treatment. Double-blind, placebo-controlled treatment with D EX was given at 2300 hr for four consecutive days (0.5, 1, 1, 1 mg, re spectively). Plasma sampling (0800 and 1600 hr) and cognitive testing (1600 hr) were performed on study days 0 (baseline), 1, and 4, and 7 d posttreatment. Repeated-measures ANOVA found a significant interactio n between study day and treatment condition for correct recall during a paragraph recall task [F(3,51)= 3.52, p = 0.02]. DEX (n = 10) in com parison to placebo (n = 9) treatment decreased correct paragraph recal l on study day 4 [F(1,17) = 5.01, p = 0.04] and study day 11 [F(1,17)= 5.82, p = 0.03], with the lowest level of performance occurring on da y 4 followed by a return toward baseline performance level by day 11. In the placebo-treated subjects, correct paragraph recall improved ove r the course of treatment, consistent with practice. No other cognitiv e measure was affected by DEX treatment, arguing against a nonspecific DEX effect on arousal or attention. Plasma cortisol concentrations we re maximally suppressed at study day 4, consistent with GC receptor bi nding by DEX.