I think the grandparent post has a point, but rather than trying to bucket disorders into whether they're a "real disease" or not, it's a lot more useful to think of a spectrum where there are some illnesses that have really clear diagnoses and mechanisms on one end, and very vague things on the other end that we hardly understand at all. Mental illnesses cluster at that more vague end.
See strep throat for a counter example (http://en.wikipedia.org/wiki/Streptococcal_pharyngitis). There's a five point diagnosis checklist that looks a bit similar to checklists you might see in the DSM for a mental illness. But there's also a gold standard test of actually culturing the cells, which is 90-95% accurate at diagnosing strep.
It may be just my ignorance, but I don't see comparably accurate tests for diagnosing mental illness. It's extremely common to see someone's diagnosis change over months or years from depression to bipolar to something else. This tends to undermine the scientific credibility of mental illness diagnosis, leading people to wonder how much diagnoses are influenced by fashions, drug companies, or the practitioner's whim.
Ninja edit: For the counter counter example, see the Rosenhan experiment (http://en.wikipedia.org/wiki/Rosenhan_experiment). There are huge confirmation bias problems in mental illness diagnosis.
I strongly agree that diagnosis is problematic. This is especially weird when people can be forced to take a medication against their will based on that diagnosis. This is really fantastically intrusive and oppressive and etc.
Another problem with diagnosis: some illnesses haw a list of (say) 9 things, and the patient must match 5 of them for 6 months. So the patient is a good match of 3 and a weak match for 2 and a poor match for the other 4. But anyone seeing that patient will read the diagnosis and will apply all the criteria to everything the patient does. "You would say that. People with your diagnosis are {manipulative}" (even if the patient was a poor match for manipulative).
See strep throat for a counter example (http://en.wikipedia.org/wiki/Streptococcal_pharyngitis). There's a five point diagnosis checklist that looks a bit similar to checklists you might see in the DSM for a mental illness. But there's also a gold standard test of actually culturing the cells, which is 90-95% accurate at diagnosing strep.
It may be just my ignorance, but I don't see comparably accurate tests for diagnosing mental illness. It's extremely common to see someone's diagnosis change over months or years from depression to bipolar to something else. This tends to undermine the scientific credibility of mental illness diagnosis, leading people to wonder how much diagnoses are influenced by fashions, drug companies, or the practitioner's whim.
Ninja edit: For the counter counter example, see the Rosenhan experiment (http://en.wikipedia.org/wiki/Rosenhan_experiment). There are huge confirmation bias problems in mental illness diagnosis.