I don't know about just dropping dead from a broken heart, sounds a bit Shakespearean to me. But times of great stress can make otherwise latent illnesses become much more active.

I, personally, would reject the notion that depression is a "clinical condition"; but then I am critical of many psychiatric practices in general.

I, personally, would reject the notion that depression is a "clinical condition"; but then I am critical of many psychiatric practices in general.

I always wondered why they could not do blood work to find out if you were suffering from a chemical imbalance.

Any basis for this? Depression has symptoms, complications, a biological basis, and a mortality rate, it causes people to present to clinical settings, and it is routinely treated by clinicians -- so what about that is not a clinical condition?

First, the phrase 'chemical imbalance' is a layperson's term. There is no disputing the abundance of literature from primary scientific sources demonstrating 1) low synaptic serotonin and norepinephrine in subjects with depression, 2) amelioration of their depression when medications are given that increase these levels, and 3) that therapy actually increases these levels comparably to drugs.

Next, the DSM committee defines disorders based on impairment PLUS the unique features of the disorder. Homosexuality does not in and of itself cause clinical impairment, so it makes no sense to categorize it as a clinical disorder. Sure, homosexuals may be at higher risk of depression and suicide, but so are people who have gotten divorced or lost their jobs, ie it's a risk factor at a population level.

Finally, you have erred by bringing up someone who has lost a family member. The DSM clearly distinguishes bereavement from major depression and other affective disorders. Bereavement is self-limited and physiologic. Some bereaved people actually respond well to a brief course of SSRI therapy, so they DO have this 'chemical imbalance'. I'd never say it like that, though, because you owe it to your patients to validate the things in their life that are weighing on them and not just blow it all off as deterministic biology.

The underlying principles of psychiatry (and the DSM) is that of a (biological) medical model, which treats mental health issues as biological diseases, when they may be better understood from a variety of viewpoints.

My point was that by providing symptoms for a "condition" (as psychiatrists did when homosexuality was in the DSM) that does not then validate that as a condition.

Whilst I accept there is evidence that some bouts of depression appear to have a biological origin, many other cases of depression may be better understood as psychological, psycho-social or a combination (etc.).

Even some mental health issues that people still regard as almost certainly being of biological origin appear (e.g. "schizophrenia") to have psycho-social factors influencing both onset and prognosis.

One of the things that bother me , is that the psychiatrist doen't always warn the patient of the possible side effects of the medicine they are prescribing.

They have to, at least common or anticipated side effects.

I hear you. But understand that medical practice is not biology. A psychiatric evaluation (and many general medical ones) needs to include a discussion about education, work, lifestyle, stressors, support systems, finances, etc. And we are in nearly constant contact with social workers, social services, financial advisors, support groups, etc to ensure that ALL relevant needs are attended to. My outpatient medical work is predominantly adolescent and pediatric HIV care, and believe me the HIV is the easy part -- it's addressing everything else in their lives that's the hard part.

That's not how the DSM is generated. The condition already exists...

The DSM is derived from clinical research studying populations of people who are impaired by a certain syndrome, and then deriving certain categories and subcategories based on their symptom complex. In this way they can broadly categorize people into have affective (mood) disorders (eg. depressive disorders, manic disorders, bipolar disorder), psychotic (thought) disorders (eg. schizophrenia), anxiety disorders, personality disorders, dissociative disorders (like fugue states, multiple personalities, etc), eating disorders, etc. I don't remember if the DSM includes dementia, delerium, etc, but either way these are other mental illnesses.

They ALWAYS have both, because we always have a brain. And any psychiatrist worth the air he breathes knows this.

Everyone knows that. Even basic biologists know that, because they know that there are measurable biological / physiological changes that happen in response to psychosocial factors.

I would dispute the notion that the condition already exists. Depression is not a concrete object, iI believe it to be more likely a construction, that is to say it is a term that has associated terms with it. One can create a term as part of psychiatric discourse and list associated "symptoms" with it; that does not mean to say that term actually "exists".

Ok, well what does exist is a subset of humans who have a common set of complaints, symptoms, and signs, that there are neurobiological correlates that occur with more specificity in this group than in the general population, and that when this population is studied certain drug therapies and cognitive therapies will improve their symptoms and life functioning compared with placebo or lack of therapy.

Even if you don't believe "the condition already exists", you cannot dispute that such people already exist. And depression or affective disorders are terms used to describe this particular constellation of findings as 1) unique and 2) distinct from the general population.

This is somewhat true, though drug therapies and cognitive therapies will:
i) Not be effective for everyone
ii) Even if it does "work", it does not work with the same efficacy in everyone, particularly antidepressants (click here or here)
iii) Thus depression does not exist as discretely as catergorical diagnostic tools such as the DSM would believe.

Yes, this is my point...they are terms (constructions) not necessarily "real" or "tangible" conditions that can be seen, or "proven".

This misses the point, though. Case report (eg anecdotal) level evidence is non-evidence in medicine. You need to study populations.

You learn about populations and about categories, and that way when evaluating an individual patient you aren't starting completely from scratch.

If I take 10,000 people who self-identify as "always sad" and 10,000 normal controls, I will find a population-level statistically meaningful difference between the study group and controls if my study questions are the described features of depression. And at that population level, if I do such a study, I will find that Prozac will work better than Penicillin, better than Procrit, better than Tylenol, better than Immodium, better than Gas-X, and better than placebo at alleviating their symptoms.

Then I'm in clinic, someone comes up to me and says they're sad all the time, they can't eat anymore, they stay up all night worrying, they feel guilty, they have no interest in anything, their relationships have all gone to hell and they can't hold down a job, and this has been going on for years. And regardless of the ontology of depression, I know that this person has a symptom complex highly similar to well-described cohorts of patients who are collectively described as having clinical depression. I also know that statistically this patient is likely to improve with antidepressant therapy, and based on other patient factors I can determine the benefit vs risk of treating him.

And given that we know that there are biological abnormalities that occur in depressed patients, it may be possible that the diagnosis of depression will be tangibly provable in the future.

By virtue of medical practices rejecting case studies as valuable evidence, or of any worth, does that mean they should be disregarded altogether?

This may depend on whether one wishes to treat people as individuals or as mere statistics.

Additionally I believe the evidence I cited with regards to antidepressants was not anecdotal.

I'm never surprised at drugs proving better than placebo, to prove that something (i.e. the drug) is better than nothing (i.e. the placebo) is not incredibly difficult.

It is possible, but is diagnosis necessarily required in the first place?

...ut when the trials are good, then a case study does not add much unless the patient in question is very different than the patients in the study (the baseline characteristics)

That's a common slogan, but that's not how statistics are used in medicine. You NEVER know which way someone will respond. But you know that other people like them usually do, almost always do, almost always don't, (whatever) respond to therapy X. That gives you a basis by which to decide if it's worth trying a therapy or not.