Components of variance in polydrug use

Forty-seven people who first attended any London drug clinic in 1971 were asked at what age they first used various drugs and how frequently each was used during every subsequent year. For age of first use, substantial significant variance components due both to drugs and subjects and to the interaction between them confirmed that certain substances were commonly tried at an earlier age, that some subjects consistently used all drugs earlier than did other subjects, and that there were numerous exceptions to these regularities. For initial frequencies of use, much of the variance was due to the interaction of drugs and subjects--meaning, in effect, that various subjects had preferences (or sources of supply) for different drugs, even though nominally all were polydrug users attending opiate clinics. However, very little variance in frequency was due to years of use or its two-way interactions with drugs or subjects. That is, once established, drugs tended to be used at about the same frequency levels over a 3-year period. The conclusions seem reasonably robust despite the comparatively small sample.