Matching strategies for drug studies of prepulse inhibition in humans.

Prepulse inhibition (PPI), a measure of sensorimotor gating, is impaired in certain neuropsychiatric disorders. Animal studies have revealed drug effects on PPI that may be relevant to understanding the biology of gating deficits in human populations. Recent efforts have examined similarities and differences in drug effects on PPI between rodents and humans. Experimental designs are needed that most effectively translate these drug studies across species. In the course of a larger set of studies of drug effects on startle in normal human subjects, we examined the potential utility of one design element that is utilized in rodent PPI drug studies: pre-testing to diminish variability across dose groups. Startle was measured during a screening session; 7-10 days later, 20 subjects were retested after consuming a placebo pill. Acoustic and tactile startle, and unimodal and cross-modal PPI, were measured in five sessions over a period of 3 hours post-placebo. There were significant and robust correlations between levels of startle magnitude and PPI during pre-testing and testing, for both left and right eyeblink measures. Comparable correlations were evident for both unimodal and cross-modal testing. Pre-testing values were most predictive of test performance early in the 3-hour test session, and predictive strength diminished or disappeared towards the end of testing. The utility of a pre-testing design could be seen clearly by comparing groups 'matched', based on pre-test data, versus groups created by alternating or random group assignments. It is concluded that pre-test designs can effectively match groups with comparable levels of startle or PPI, and thereby diminish between-group variability in human PPI drug studies. For studies using repeated testing to assess drug time course, the predictive value of pre-testing is greatest in early test sessions.