To Include or Not Include Comorbidities Posted on October 19, 2007 Husseini K. Manji (bio) discusses why he generally chooses not to include comorbidities in research.

From a research design methodology, I really think it depends on the type of research one is doing. If one is doing research design that, the term is sometimes used, pathophysiology, to try and understand what’s wrong in the illness. Then, I think, my group errs on the side of trying to study the pure disorder first, to try and see what’s going on in bipolar disorder when you don’t have the comorbid alcohol abuse or comorbid other illnesses, anxiety disorders are common, etc., and then if we find something on a brain imaging or on a genetic basis, then we try and see does this also hold true when you have the comorbidity or, in fact, you have more of the problems if you have the comorbidity, etc.

And the reason to do that is sometimes if your population isn’t very clean, and I use the term ‘clean’ to mean without comorbidities, it’s difficult to know, am I seeing this finding because of the bipolar, because of the substance abuse, a little bit of each, etc.? And so sometimes when you’re trying to find new things, you need to be a little bit more restrictive, and then, of course, you have to try and see does it hold up in the general population.

The flip side of that is that if you’re trying to come up with, for example, a new treatment, then it might make more sense to try and see what works in the general bipolar population, because it’s not necessarily too useful to find a treatment that only works in bipolar disorder if you don’t have any comorbidities, etc. So for those kinds of studies, what sometimes the term used is “real world studies,” I think, in fact, including the comorbidities makes more sense, because that’s what the majority of patients have.

« Back to Article