Changing Diagnostic Criteria Did Not Expand Cannabis Dependence Uniformly in a Historical Comparison
A historical comparison of DSM-III and DSM-III-R diagnoses found that neither system consistently predominated for cannabis dependence. Changes that expanded diagnoses for some other substances did not necessarily produce the same pattern for cannabis.
The study compared classifications in the same setting
Recent admissions to drug-treatment centers were interviewed with an instrument covering criteria from both systems. Investigators examined agreement, overlap and positive agreement by substance. These measures assess correspondence between classification systems, rather than the biological presence of a drug or impairment at a particular time.
Results differed across substances
The abstract reports a wider dependence classification under DSM-III-R for alcohol, tobacco and amphetamines, but no predominating system for cannabis, opioids or sedative-related categories. A general statement that the revision simply increased every substance diagnosis would therefore be unsupported.
Historical labels cannot be substituted for current criteria
This was a comparison of older diagnostic systems in a treatment-seeking population. It does not establish the prevalence of a current cannabis diagnosis in the general public. Nor can a toxicology result alone determine which interview-based criteria a person met. Drug detection and longitudinal clinical assessment address different evidence.
Interpreting the evidence
When comparing cannabis diagnoses across studies, retain the criteria edition, instrument and population. Apparent changes in diagnostic counts can reflect how a category was defined as well as changes in the people assessed.
Related reading: separate functional and structural endpoints in cannabinoid research.
Research reference
Cottler LB; Helzer JE; Mager D; et al. (1991). Agreement between DSM-III and III-R substance use disorders. Drug and alcohol dependence. doi:10.1016/0376-8716(91)90018-t. Evidence basis: report abstract.
