My first thumbing through the DSM-III in 1980 landed me in Major Depressive Disorder [MDD], and it’s where I’ve remained. I’ve written about it enough here to thoroughly earn my 1boringoldman moniker [see a mistake…, further thoughts on the mistake…, and yet another mistake… for just one version]. In their zeal to bring psychiatric classification more into the medical realm, they eliminated the most medical of all psychiatric diagnoses – Melancholic Depression. As I’ve said before here often, in a former time the word depression had different connotations than it does today. It referred to a felt emotion, something all of us are capable of feeling – a biological given. A major clinical distinction was depression, the emotion, and Depression, the illness known clinically as Melancholic Depression, which is both experienced and observed as something very different from depression [see melancholia…]. Several years back, Dr. Bernard Carroll put Melancholia to verse [right][see Bringing Back Melancholia]. On the road to the DSM-5, a Who’s Who from the ranks of psychiatry lobbied in vain to have this diagnosis returned to the official classification [Note: That list of authors included Dr. Robert Spitzer himself, who was the person who had eliminated it some thirty years before].by Gordon Parker, M.D.; Max Fink, M.D.; Edward Shorter, Ph.D.; Michael Alan Taylor, M.D.; Hagop Akiskal, M.D.; German Berrios, M.D.; Tom Bolwig, M.D.; Walter A. Brown, M.D.; Bernard Carroll, M.B.B.S.; David Healy, M.D.; Donald F. Klein, M.D.; Athanasios Koukopoulos, M.D.; Robert Michels, M.D.; Joel Paris, M.D.; Robert T. Rubin, M.D.; Robert Spitzer, M.D.; and Conrad Swartz, M.D.American Journal of Psychiatry 2010 167:745-747.
Melancholia, a syndrome with a long history and distinctly specific psychopathological features, is inadequately differentiated from major depression by the DSM-IV specifier. It is neglected in clinical assessment [e.g., in STAR*D] and treatment selection [e.g., in the Texas Medication Algorithm Project]. Nevertheless, it possesses a distinctive biological homogeneity in clinical experience and laboratory test markers, and it is differentially responsive to specific treatment interventions. It therefore deserves recognition as a separate identifiable mood disorder.Melancholia is a lifetime diagnosis, typically with recurrent episodes. Within the present classification it is frequently seen in severely ill patients with major depression and with bipolar disorder. Melancholia’s features cluster with greater consistency than the broad heterogeneity of the disorders and conditions included in major depression and bipolar disorder. The melancholia diagnosis has superior predictive validity for prognosis and treatment, and it represents a more homogeneous category for research study. We therefore advocate that melancholia be positioned as a distinct, identifiable and specifically treatable affective syndrome in the DSM-5 classification.
I recently read a remarkable paper by the lead author above, Dr. Gordon Parker of The Black Dog Institute in New South Wales Australia, that came at this distinction in a unique way:
by Parker G, Paterson A, and Hadzi-Pavlovic D.Acta Psychiatrica Scandanavia 2015 Jan 6. [Epub ahead of print]
OBJECTIVE: We sought to determine whether putative depressive diseases could be differentiated categorically from clinical depressive disorders and non-clinical mood states.METHOD: We interviewed volunteers who reported or denied any lifetime depressive mood state and analyzed data from the former group reporting on their ‘most severe’ depressive episode. We employed latent class analysis [LCA] to determine whether a two-class solution was supported and the contribution of individual variables to class allocations.RESULTS: All variables were significant predictors of class allocation. LCA-assigned Class I participants reported more depressive symptoms, had more distressing episodes and more lasting consequences, were more likely to view their depression as ‘like a disease’, and as being both disproportionately more severe and persistent in relation to any antecedent stressor. Validation involved comparison of LCA assignment with DSM-IV diagnosis for their most severe depressive episode. Of those assigned to Class I, 89% had a DSM diagnosis of melancholic, psychotic or bipolar depression. Class II had all those failing to meet criteria for a depressive episode and the majority of those with a non-melancholic depressive condition.CONCLUSION: Despite not including individual depressive symptoms, study variables strongly differentiated putative depressive diseases from a composite of clinical depressive conditions and subclinical depressive states.
The point here is that those unique symptoms we’ve always seen as defining Melancholia moved with this study from prose and poetry to an objective instrument. And complex cluster analysis of the results of that instrument in a mixed cohort of subjects with and without a previous depressive episode dramatically separated them into two relatively homogeneous groups – melancholic versus something else. Further, cross checking those clusters with conventional diagnostic criteria [DSM-IV] was a surprisingly good fit. I hope this paper becomes freely available; gets presented in more detail elsewhere; and is scheduled for replication trials forthwith. So far, Melancholic depression separates from the other depressions as a unique entity as cleanly as one could hope from the technology available for such an analysis. Melancholia is a thing of its own just like it has always been – not a feature of something else which never was. Maybe we can now pick up this thread that went on political sabbatical in 1980, and figure out what Melancholia is after all…
I would love to see new large scale antidepressant trials redone on the group with melancholia, five groups, placebo, SSRI, ,TCMS, TCA, MAOI. I would predict a higher response rate for the last two categories and a much lower placebo response. By weeding out self-limiting conditions, we can start making sense of thirty five years of garbage.
What will actually happen though is that studies will continued to be done using DSM or the PHQ-9 and placebo responses will continue to grow.
James
Tolly agree!
James, here is an early look at that question:
Roose SP, Glassman AH, Attia E, Woodring S. Comparative efficacy of selective serotonin reuptake inhibitors and tricyclics in the treatment of melancholia. Am J Psychiatry. 1994 Dec;151(12):1735-9.
PubMed ID: 7977878. The workhorse tricyclic antidepressant drug nortriptyline beat the SSRI drug fluoxetine hands down.
I watched the movie Melancholia. You said it was representative, and I believe you. It’s a crying shame that that is not in the DSM anymore, and that there isn’t research focusing on that condition, at all times. It was a very vivid portrayal of a truly disabling condition that William Styron (hope I got the name right) also described such an overwhelming and all-consuming depression. The Major Depression classification eclipses this horrid illness which is an injustice every bit as inexcusable as giving people medication for being appropriately miserable.
FREE journal article about what a crappy job the FDA is doing with drug testing.
http://archinte.jamanetwork.com/article.aspx?articleID=2109855
Six groups– add an ECT trial to seal the deal.
Actually seven groups–add a ketamine trial.
Thank you Dr. Carroll for that link.
Interesting to see those results in elderly with cardiac problems because you rarely see a TCA anymore in elderly with cardiac problems. Sometimes Desyrel.
Is there anything similar for younger patients?
If TCAs beat SSRIs in younger patients under 25, there really is no reason to use SSRIs at all in younger patients over TCAs unless demonstrated intolerance to TCAs or MAOIs. Dry mouth is annoying but far less important than suicide risk
All of this is consistent with your earlier numbers on NNT TCA vs. SSRI/SNRI.
http://www.ncbi.nlm.nih.gov/pubmed/14632962
Yikes
RE: and figure out what Melancholia is after all…
I think this extraordinary piece of scientific synthesis may help quite a bit with this: The Depths: The Evolutionary Origins of the Depression Epidemic. It “offers a bold new account of why depression endures” – really worth your time to have a close look at, I think, Mickey and all. Interesting to see Kindle and also Audio editions. http://www.amazon.com/The-Depths-Evolutionary-Depression-Epidemic/dp/0465022219
I am not familiar with the science of melancholia, vaguely remembering that there were some hormone biomarkers.
My questions are:
(1.) Is there a good way to distinguish melancholic unipolar depression from a severe bipolar depression (other than history)?
(2.) Do the tricylics risk inducing mania in those vulnerable to bipolar disorder?
The problem with ECT in the trial is that it’s not a first line treatment.
The problem with ketamine is that it’s still experimental and probably only lasts for a couple of weeks.
This has gotten me to thinking that the whole algorithm of depression treatment needs to be revamped.
Mild to moderate=exercise, sleep hygeine and psychotherapy.
Moderate to severe=TCA and psychotherapy.
“Atypical”or first line failure=MAOI and psychotherapy.
I’m increasingly wondering what the role is for SSRI unless someone is already on it and its working. SNRI might have some carve out in pain/depression syndromes or perimenopausal depression/anxiety. SARI will still have a role for sleep and depression and some anxiety disorders.
I have seen a lot of weight gain with TCAs and MAOIs, more so than with SSRIs.
To EastCoaster’s questions: (1) People have looked hard for features that might be a tip-off for bipolar depression in patients with what appears to be a unipolar depressive disorder. Here is one such attempt: nothing really solid was found though there were trends pointing to psychotic features, diurnal mood variation, and hypersomnia as predictive features.
(2) As for antidepressant drug-induced switch into mania, check here and here. In short term trials with unipolar patients the switch rate is less than 1%, while in bipolar patients it is around 11% with tricyclic antidepressants, 4% with SSRIs and 4% with placebo. In longer term observations the rate of switch in unipolar cases seems to be higher at around 8% over 2 years. The rate is higher in youth than in adults. No clear predictors of the switch have been identified.
Dr. O’Brien,
I like your treatments for mild to moderate depression. I would also propose for all levels of depression that patients be screened for possible sleep disorders such as UARS and sleep apnea. I wish I had a nickle for all the folks in apnea boards who spent years on psych meds only to find they had sleep apnea. So that is why I may seem anal retentive about this. 🙂
Dr. Arpaia, I gained 30 pounds on Prozac and greatly feared that if I hadn’t switched to St. Johns Wort, the weight gain would have continued. It was extremely frightening. Thankfully, once I made the switch, the weight came off effortlessly.
With melancholia, weight gain is often desired. Parnate beats Nardil and TCAs if this is a problem. Probably underused in overweight patients with depression.
Has anyone here used EmSam patch? This has been out for a while I haven’t spoken to any psychiatrist who has tried it yet. The cost is a big issue.
Thanks, Dr. Carroll. To me the lack of a biomarker for bipolar disorder concerns me, because the melancholia population seems too heterogeneous. But obviously, it’s way more precise than the current system.
To EastCoaster: Yes, even the restricted concept of melancholia carved out from DSM-5 major depressive disorder is heterogeneous. I frame it as 3 subgroups under the general umbrella term melancholia. One group have late onset, a generally negative family history, and might nowadays be viewed as vascular depressions (the old involutional melancholia). Then there are those with delusional melancholic depressions, who present somewhat different management issues. Finally there are the early onset, family history positive patients with melancholic episodes. At least all three types respond to classic antidepressants and to ECT if necessary.
By the way, with reference to switching, it is worth noting that the most common psychiatric disorder in the extended pedigrees of bipolar patients is unipolar depression. The family tree of Virginia Woolf is a good illustration of that.
This meta analysis http://www.ncbi.nlm.nih.gov/pubmed/22298630 tries to solve the question about the comparative efficacy of SSRIs vs TCAs, and the findings are equivocal, though they did not test specifically the melancholic/non-melancholic sub-groups.
It is no accident that there are so few studies of the new antidepressant drugs in melancholia. As I wrote here, “In pursuit of mass marketing, pharmaceutical corporations settled for weak evidence of efficacy in undifferentiated major depression – they had no financial incentive to pursue the study of depressive subtypes. Thus, the populations tested today in drug registration trials differ in important ways from the clinical research populations in whom antidepressant drug efficacy was first established. The result is an epistemologic quagmire about the efficacy of today’s most widely used antidepressant drugs.”
Here’s an interpretation to get me in trouble yet again, after I foolishly shared it at a pharmacology lecture my first year of residency about 25 years ago, note before the SSRIs really took off as alternatives to the TCAs and MAOIs:
“If the placebo response in most antidepressant trials was about 35% on average, and the response rate to antidepressants was about 65-70%, then why do we say antidepressants are twice as effective? To me, they just impact on about a third of patients, and if we really respect the Hippocratic Principle of “first do no harm”, why are we pushing meds when an equal amount of people will show improvement without medications that can have fairly impairing side effects?”
Got a lot of stares and looks of astonishment from classmates and the lecturer. Ironically, to me, it still holds true today, as I think honest and responsible psychotherapy will treat as many patients with depression as those who just take medications. In fact, we all know the combo of the two is the most effective for those who have real clinical depression, so, after forty years or more of hearing about placebo responses, why are we still acting like that is simply a phenomenon, and not part of the life cycle of depression?
The academic commenters here will probably be staring at the screen right now, the clinical ones just muttering, and the people who have taken these meds, well, their responses will be divided between “no, meds work”, and the other near half of “exactly our point”.
Not that I do not prescribe nor do I have no faith in medication, but, why are truths still ignored to this day?
Think about it.
And today the placebo rates in trials is like 50%….what does that say about the conditions that are being selected….self limited formes frustes basically…translation…actually adjustment disorders that were going to get better anyway….hence exercise and sleep hygiene and psychotherapy makes sense…
A 35% placebo response rate today would be an improvement on what is happening today….we have written here about the explosion in placebo response and how this is killing research….and as Dr. Carroll said making true comparisons impossible. It’s like trying to figure out if Barry Bonds 2001 was a better slugger than Babe Ruth 1927 or Roger Maris 1961 when you suspect he’s cheating on roids…which turned out to be the case. You can roid out your study results by adding a lot of people who will get better no matter what…
BTW I want to point out that psychiatry is not the only field in which diagnostic and treatment inflation is creating a huge overtreatment problem. Mild hypertension and hypercholesterolemia without history of heart disease are major areas of controversy right now. If you go to the NNT site, it is not recc to treat these with ACE inhibitors and statins, correspondingly. But this has really been the standard in clinics for years. There was a major article in WaPo yesterday about the government backing off its dietary cholesterol reccs.
I guess since high cholesterol was associated with heart disease and strokes, it made sense to take statins to reduce cholesterol (more specifically LDL). But as it turns out, even though that association is correct, the benefits of reducing cholesterol through medication are not that great, and not even as good as going on a Mediterranean diet.
And so it is with mild depression. Turns out nonprescription methods work better for that as well.
This is all a bit unsettling but not surprising I guess. Look how long it took to figure out the role of fat/protein/carbs in dieting.
Educate me. If melancholia is the most severe form of depression, life threatening and functionally-debilitating in effect, and if ECT has been shown to be the most effective treatment for this illness, then why on earth isn’t ECT a first line treatment for this horrible malady? Why screw around with ineffective or partially effective medication trials at all?
Tom, I think there’s plenty of smart money out there that likes ECT for depression, but it’s really invasive, I don’t think anyone really knows how it works, and I’m not sure the side effect profile is well understood, either. The other issue is treatment failure– man, when it does fail, those clients are REALLY depressed, or at least that’s my understanding.
The other lifestyle modification I would add to Dr. O’Brien’s list would be alcohol and cannabis cessation, at least for most patients in the moderate to severe category. Both these substances sometimes seem to have some positive effects in moderation for folks in their mid 20s to mid 40s, and cannabis certainly seems to have a place in the arsenal of pain control and anti-inflammatory agents, but in my very limited experience, I have not seen a single client who didn’t recognize significant improvement when they stopped these two. Particularly patients in their 50s– I’ve seen multi-year dysthymic episodes lift completely with cannabis cessation.
I admit to some bias because my own dysthymia or depressive episodes really cleared up quickly when I stopped. Severity was the same, but the duration dropped from weeks or months to 36 hours max.
And yes, I know there is no such thing as a dysthymic episode, but after reading this blog, I’m really not sure what to call it. I don’t know whether it was melancholic depression or not.