A Rating Scale for Depressive Disorders

Abstract

This rating scale has been devised for use with patients already diagnosed as suFfering from a depressive disorder. It is applicable to all types of depressive disorders, provided that it can be accepted that the patient's primary disturbance is that of a change in mood. The rating scale consists of 17 items, some of which have 5 levels (absent, mild, moderate, marked, severe), and others only 3 (absent, doubtful or trivial, present). A total score on the scale gives a satisfactory measure of the severiry of the illness. The preliminary report (1) has now been supplemented by further cases. The standardization population now includes 152 men of known characteristics. The means and standard deviations of the items and the intercorrelation matrix are available.' Work with the scale has proceeded along two main directions. The first is concerned with an analysis of the scale and ics use in classification of the depressive disorders. The second is concerned with the use of the scale for quantification and its application to the prediction of outcome of treatment in these disorders. The scale itself has been facror analysed by the method of principal components and 6 factors extracted. As is to be expected with a rating scale used on a relatively homogeneous population, the intercorrelations between variables are low in consequence, the latent roots are small and diminish slowly, but the first 6 are the only ones that have a value greater than unity. The first factor may be regarded as a general factor of depression, and facror scores correlate extremely highly with raw total scores on the scale. The second factor is a bipolar factor which counterbalances the typical symproms of classical retarded depression at the positive end, against the typical symptoms of agitated depression at the negative end. It can, therefore, be regarded as a retarded-vs-agitated facror. The other factors are nor so clearly definable and it is unlikely that the 5th and 6th factors are sufficiently stable to be worth considering. Factor scores for the population, in T form, have been computed and it has been found that all of them show normal distributions except the second, which is non-normal (p = .05). This bimodal distribution provides the first objective evidence, based on symptoms, that retarded and agitated depressions are distinct disorders and not merely extremes of a distribution. A Varimax rotation of the principal factors has been made, and the first two Varimax factors can be seen quire clearly to represent retarded depression and agitated depression; the other factors do not appear to correspond with any of the traditional clinical syndromes. Further work is being done on the Varimax factors. In the case of those patients who have been treated and followed up, the items of the rating scales have been correlated with the outcome of treatment. Correlations are low and in some cases there is a statistically significant non-linear relationship. The relation benveen this and the possibiliry of dichotomising the depressive disorders is now being examined.