Depressive patients differ strongly in their presented symptom patterns, in how they respond to treatment, and in how their depression develops over time. These differences are poorly understood and a likely reason for the lack of advances in depression research and clinical treatment.
In his dissertation, Rob Wanders investigates depression and its measurement using a data-driven approach that is not limited to current diagnostic schemes. In several large patient and population studies, advanced statistical models were used to investigate both similarities and inconsistencies in depressive symptom patterns across patients.

The results give new insights in individual differences of depression by detecting groups of patients with similar symptom patterns. Only groups with mixed presentations of depressive and anxiety symptoms were found. In addition, a large part of the population does not meet diagnostic criteria, yet experiences symptoms of depression and anxiety with associated disability. Data-driven models were also useful to detect patients with patterns that deviated strongly from the other symptom patterns in the data. Most of these patients are complex cases, with multiple disorders occurring together in the presence of depression. Data-driven techniques have merit for clinicians as they can offer additional patient information, and can provide signals for different needs of care.
Findings of these studies showed ample empirical evidence for many aspects of current diagnostic systems. Diagnostic nets should be cast wider in research and clinical practice to allow for a clearer focus of efforts to decrease the burden and costs of depression and anxiety in the population.