Statistical relationships are not diagnoses
Scientists have applied powerful statistical techniques to try to understand whether people's experiences of mental health problems do cluster together in the way predicted by the diagnostic approach. The central idea of diagnosis is that particular psychological problems cluster together. Diagnosis depends on a particular 'disorder' or 'illness' having shared characteristics that make it distinct from other 'disorders' or 'illnesses'. But in general the results of this approach have not been supportive of the diagnostic model. There do appear to be patterns and relationships between problems (of which more later) but these relationships tend to cut across diagnostic categories.
It is important to note that this applies - perhaps more powerfully - when we explore the role of biological factors. It was with some fanfare that researchers announced13 that they had identified some genetic factors associated with a range of common mental health problems. It is important, and significant, work. But these genetic characteristics, the associated biochemical pathways and the psychological processes that affect and are affected by them, are not specific to particular diagnoses. Quite the reverse, the biological and psychological pathways quite clearly cut across these diagnostic categories.