Differences in drug intake levels (high versus low takers) do not necessarily imply distinct drug user types: insights from a new cluster-based model

Diego M Castaneda,Martin Job
DOI: https://doi.org/10.1101/2024.07.15.603634
2024-07-18
Abstract:Background: Classifying psychostimulant users as high and low responders based on median split of drug intake levels has face-validity: these appear to be different types of drug users. However, because psychostimulant intake levels a) are defined by an inverted U-shaped dose response (IUDR) curve, and b) do not necessarily imply motivation for the drug, it is unclear that median split-designated high and low drug responders represent different drug user types. Aims: To determine if median split-designated groups of high and low drug takers represent distinct groups when subjected to a new cluster-based model. Methods: Male Sprague Dawley rats (n = 11) self-administered cocaine doses (0.00, 0.01, 0.03, 0.1, 0.3, 0.56 and 1.00 mg/kg/infusion) to reveal the IUDR curve per individual. We derived six variables defining the structure of the IUDR curve (amplitude, mean, width, and area under the curve: AUC) and the IUDR-derived economic demand curve (consumption at zero price or Q0 and the motivation for drug or α). We compared median split and clustering of all variables (cocaine dose, IUDR/demand curves) obtained. Results: Median split of individual cocaine doses and IUDR curve-derived variables identified high versus low responders, but these groups were inconsistent with regards to group composition. Clustering of all cocaine doses revealed one cluster. Clustering of IUDR curve-derived variables revealed one cluster. Global clustering of all cocaine doses and all IUDR curve-derived variables revealed only one cluster. Conclusions: High and low drug takers do not necessarily represent distinct drug user types.
Neuroscience
What problem does this paper attempt to address?