Abstract
Objective: Harmonizing different depression severity scales often requires creation of categorical variables that may decrease the sensitivity of the measure. Our aim was to compare the associations between categorical and continuous and harmonized measures of depression and global functioning in a large dataset of older age patients with bipolar disorder (OABD).
Method: In the Global Aging & Geriatric Experiments in Bipolar Disorder Database (GAGE-BD) the 17-item Hamilton Depression scale (HAM-D), Montgomery Asberg Depression Rating Scale (MADRS) or the Center for Epidemiological Studies Depression scales (CES-D) was used to assess current depressive symptoms, while the Global Assessment of Functioning (GAF) assessed functional status. Data were harmonized from 8 OABD studies (n = 582). In each subsample, the relationship of depression severity as a continuous and categorical measure was compared to GAF. In the total sample, harmonized ordinal depression categories were compared to GAF.
Results: Effect size and variance explained by the model for the categorical measure in the total sample was higher than both the categorical and continuous measure in the CES-D subsample, higher than the categorical but lower than the continuous measure in the HAM-D subsample, and lower than both the categorical and continuous measures in the MADRS subsample.
Limitations: All included studies have different inclusion and exclusion criteria, study designs, and differ in aspects of sociodemographic variables.
Conclusions: Associations were only slightly larger for the continuous vs categorical measures of depression scales. Harmonizing different depression scales into ordinal categories for analyses is feasible without losing statistical power.