Background We critically reviewed the arguments of the symptom-oriente
d researchers who propose to replace syndromes and diagnostic categori
es with symptoms as units of analysis in psychiatric research. Method
Three central arguments were examined: (a) current diagnostic categori
es lack reliability and validity; (b) using diagnostic categories lead
s to misclassification and confounding; and (c) symptom - oriented the
ories are clearer, easier to test, and more likely to lead to an expla
nation of psychopathology. These arguments are based on three assumpti
ons respectively: (a) symptoms have higher reliability and validity; (
b) underlying pathological processes are symptom-specific; and (c) elu
cidation of the process of symptom development will lead to (and must
precede) the discovery of the causes of syndromes. Results We found li
ttle evidence supporting these assumptions and arguments based on them
. Conclusion There are no clear advantages in replacing syndromes with
symptoms as units of analysis for psychiatric research.