Psychometric analysis of the Melancholia Scale in trials with non-pharmacological augmentation of patients with therapy-resistant depression.

OBJECTIVE

The Melancholia Scale (MES) consists of the psychic core items of the Hamilton Depression Scale (HAM-D6) (depressed mood, interests, psychic anxiety, general somatic, guilt feelings, and psychomotor retardation) and the neuropsychiatric items of the Cronholm-Ottossen Depression Scale. Patients resistant to anti-depressant medication (therapy-resistant depression) have participated in our trials with non-pharmacological augmentation. On the basis of these trials, we have evaluated to what extent the neuropsychiatric subscale of the MES (concentration difficulties, fatigability, emotional introversion, sleep problems, and decreased verbal communication) is a measure of severity of apathia when compared with the HAM-D6 subscale of the MES.

METHODS

We have focused on rating sessions at baseline (week 0) and after 2 and 4 weeks of therapy in four clinical trials on therapy-resistant depression with the following augmentations: electroconvulsive therapy, bright light therapy, transcranial magnetic stimulation or pulsed electromagnetic fields, and wake therapy. The item response theory model constructed by Mokken has been used as the psychometric validation of unidimensionality. For the numerical evaluation of transferability, we have tested item ranks across the rating weeks.

RESULTS

In the Mokken analysis, the coefficient of homogeneity was above 0.40 for both the HAM-D subscale and the apathia subscale at week 4. The numerical transferability across the weeks was statistically significant (p < 0.05) for both subscales.

CONCLUSION

The apathia subscale is a unidimensional scale with acceptable transferability for the measurement of treatment-resistant symptoms, analogue to the psychic core subscale (HAM-D6).