Background: Despite its importance, no distinction between moderate and sev
ere depression using the Montgomery-Asberg Depression Rating Scale (MADRS)
based on a direct comparison with the Hamilton Depression Rating Scale (HAM
D-17) is available. Methods: HAMD-17 and MADRS ratings from N = 40 at least
moderately depressed inpatients with major depression (DSM-III-R) were ana
lyzed. Linear and non-parametric correlations were computed and a MADRS cut
-off score for severe depression using an HAMD-17 score of at least 28 poin
ts as reference was estimated. Results: HAMD-17 and MADRS mean scores were
24.6+/-4.3 and 32.6+/-5.0 points, respectively. Linear correlation of both
scores was r = 0.70 (P < 0.0005). A MADRS cut-off score of at least 35 poin
ts was estimated to separate 'moderate' from 'severe' depression correspond
ing to a HAMD-17 cut-off of 28 points with sufficient sensitivity and speci
ficity. Limitations: The sample size was limited and no observer ratings di
rectly assessing the severity of depression were used. Conclusions: The pre
liminary findings are in line with previous findings and suggest a cut-off
score of 35 points to separate moderate from severe depression with the MAD
RS accepting an HAMD-17 score of greater than or equal to 28 point as refer
ence. Further studies on this issue are warranted. (C) 2000 Elsevier Scienc
e B.V. All rights reserved.