Quantitative evaluation of cue-induced reinstatement model for evidence-based experimental optimization.

Cue-induced reinstatement is a widely used model for investigating relapse of reward-seeking behavior with high face validity in relation to clinical observations. Yet, face validity is not sufficient to evaluate an animal model, and quantitative, evidence-based analysis is required to estimate the ultimate applicability of this paradigm. Furthermore, such analysis would allow an accurate and reproducible design of future experiments. Here, we conducted meta-analysis and cluster analysis to characterize the impact of cue type (visual, auditory, olfactory or combinations thereof), intensity (e.g. light frequency, sound volume and odor concentration), reward type (e.g. different drugs of abuse, sucrose and food pellets) and model parameters (e.g. housing condition, age, gender and strain of animals) on reinstatement levels. We selected 184 publications for meta-analysis based on inclusion criteria with a total number of 3889 rats. Our analysis suggested that the exact level of reinstatement depends on neither cue type, nor intensity nor on the type of reward. While all cues induced reinstatement to reward-seeking behavior, it is the model parameters, in particular, the housing conditions, age and strain, that defined the final reinstatement levels. In particular, single-housed, adolescent, Wistar or Lister Hooded rats showed significantly higher reinstatement than adult, Sparague-Dawley rats housed in groups. Our findings suggest that model parameters (for example, single housing) evoke stress-induced behaviors that affect reinstatement more than cue/reward factors.