Rationale: The fear-potentiated startle paradigm [increased startle in the presence of a conditioned fear stimulus (CS)] has become increasingly popular as a tool for evaluating the potential efficacy of putative anxiolytic compounds. However, when the tested compounds also influence baseline startle, it is unclear how comparisons with control groups can best be made.
Objective: To evaluate the validity of absolute difference (startle amplitude on CS minus non-CS test trials) vs. proportional increase (the absolute difference score divided by startle amplitude on non-CS test trials) scoring methods.
Methods: The effect on proportional increase and absolute difference scores of baseline shifts that occur with or without concomitant increases in fear was evaluated in rats. A reliable measure should yield similar scores across shifting baselines, provided that fear levels remain constant.
Results: Preexisting baseline differences, and those brought about by different startle-eliciting noise burst intensities, by strychnine injections, or by CRH infusions, each increased absolute difference scores without markedly influencing proportional change scores. These baseline differences were not associated with different fear levels. Increases in baseline startle brought about by unsignaled footshocks or by a second CS - increases which are associated with increased fear - partially occluded additional CS-induced increases using either measure.
Conclusions: Across different baselines, CS-elicited fear is most accurately reflected in proportional change scores. Under certain conditions saturation effects may interfere with an accurate assessment using either measure. However, these same saturation effects may provide opportunities to explore the neural circuitry of fear and anxiety in novel ways.