The negative component in schizophrenia.

Many conflicts in the literature about negative symptoms arise because different authors use different definitions of negative symptoms. If narrow definitions are used, it appears that drugs have little effect against negative symptoms, although those who use broader definitions tend to conclude that negative symptoms do respond to drugs. Another complication is that negative symptoms may be secondary to other causes, such as psychotic symptoms, the side effects of drugs, depression and understimulation as a result of hospitalisation. Attempting to measure drug effects on negative symptoms can be complicated by such factors as different co-medication in the test and control groups and different pharmacological profiles of test and standard drugs, in particular in terms of extrapyramidal side effects or antidepressant effects. Several approaches to dissecting the direct effect of drugs from their indirect effects on negative symptoms have included analyses of covariance, regression analyses and path analyses. Applying path analysis to the data from the North American clinical trial of risperidone suggests that a significant component of the advantageous effect of risperidone on negative symptoms, compared with haloperidol, is due to a direct effect that cannot be accounted for by the effect of risperidone on positive symptoms and its superior side-effect profile. Although it is possible to analyse studies retrospectively in this way, the best approach is probably to design future clinical trials so that the effects of drugs on negative symptoms are more easily determined than has been the case in the past.