Tuesday, January 22, 2013

I am not the only one that thinks "Blue Monday" is a hoax.

This ridiculous attempt at pseudoscience is getting a rough ride, and it is well deserved. Some health reporters are able to see through this, e.g. at CTV. Other bloggers have also been critical, for example, here. Hopefully, this concept will just disappear over time.

Monday, January 21, 2013

The "Blue Monday" hoax

A strange phenomenon seems to be gaining people's attention - the idea that a day in the middle of January is the gloomiest day of the year. Here is an example of a newspaper article about this. The identification of this date is purportedly based on a calculation - which on the surface seems completely bogus. The parameter includes equations such as "weather" and "need for action" and "debt" - things that are not directly measurable and certainly things that one suspects cannot be added or multiplied together to identify a date. There appears to be no discussion anywhere about whatever units would factor into these calculations, nor any theoretical or empirical validation of this equation. As has been pointed out by others (e.g. click here), this appears to be a pseudoscientific hoax that has really taken off. A summary of silly press releases is even available.

Wednesday, January 16, 2013

Diathesis-Stress Models of Major Depression Epidemiology

Several of the postings in this blog have been concerned with the limitations of the current diagnostic conception of major depression. The concerns expressed are not idiosyncratic and have been expressed by many others. The main problems arise from the need (when creating a diagnostic definition, which is a named, or nominal, diagnostic category) to apply some sort of yes/no distinction to emotional states - which are by nature fluid, variable and non-discrete. In DSM, a threshold-based approach is applied whereby people having lots of symptoms, for a long-time and in association with problems (e.g. distress, dysfunction) are considered to have a major depressive episode. There is obvious clinical value to being able to make a diagnosis, but there is a troublesome arbitrariness to imposing a nominal categorization onto what is quite clearly a dimension of experience.

A question that I've had for a long time is whether an alternative conception could be developed, so I've begun attempting to work on this problem. The approach that I've taken is a "diathesis-stress" model approach, which treats each person as having a diathesis level (vulnerability) and as being exposed at any point in time to a stress level, and depressive symptoms are seen as arising from an interaction between diathesis and stress. This week, a description of this work has been published in BMC Psychiatry. A link to the paper may be found here....

The paper has links to some models, which were set up using the free software NetLogo - and are available to anyone that might want to play with them through links contained in the references in the paper (35, 36 and 37).  The first model, which you can link to here...

...illustrates how the model works - a person is simulated by a yellow triangle, which moves over a landscape characterized by different levels of stress (represented by shaded coloring on the model interface), as the "agent" encounters higher levels of stress, there is an interaction between their stress and diathesis, which results in emergence of some level of depressive symptoms. Lots of detail about this and more sophisticated models are available in the paper.