The goal of depression is to drive least valuable members of the pack to suicide. This has a number of potential benefits:

1. First and foremost the pack can only be as numerous as the resources allow, if the reproduction rate is higher then what can be sustained then it is a sound idea to sacrifice the least important members.

2. The symptoms of depression – severance of social contacts, no food, no procreation, no will to act, suicidal thoughts, are consistent with the individual wanting to die. Such an individual would naturally make an easy target for predators and in that way could offer some protection for the more valuable members.

This is much more beneficial then an outright killing of unwanted members of the pack since the latter would promote internal conflicts and infighting as the individuals closer to the singled out target would try to defend him or retaliate. By having the individual commit suicide this problem is averted. What’s more as the individual get’s depressed he actively tries to sever the ties with those who were close to him before and it can be interpreted as an attempt to limit their suffering when he will eventually die.

This hypothesis would also explain why social exclusion can often lead to depression since if this hypothesis is correct it would be the exclusion by individuals perceived as authoritative and influential in the group to which one aspires that would be the signal inducing depression.

So to sum it up the benefits are:
1. It is beneficial for the pack to have a way to get rid of unwanted individuals in a non-violent manner, the reasons may be lack of resources, internal conflicts, removal of individuals who’s behavior puts the group in danger, etc.
2. It may be beneficial for the pack to single out certain individuals as a sacrifice for predators to make other more valuable individuals safer.
3. It could also serve as a mechanism protecting other members from diseases if the diseased individuals where to get depressed and leave the pack to die but this poses the problem of why healthy individuals suffer from depression.

The point of my post wasn’t to advocate for any particular evolutionary account of depression, it was to point out that the rumination hypothesis discussed in the link article seems (to me) less plausible than many others. So I don’t think we disagree on that point, and I’m not sure how you got that impression. As for your ATP conjecture, I don’t find it particularly convincing, but I think it’s an open empirical question. I think there are any number of possible explanations for depression that remain viable, and time will tell.

First, if “rumination” (i.e. long term cognitive problem solving and planning) were the primary benefit, then animals that don’t do long term cognitive problem solving and planning (i.e. just about all animals) should not exhibit depressive symptoms. Rumination might be a human adaption to use time spent in a depressive state productively, but that can’t be the feature that depression evolved to achieve because animals that don’t plan still can get depressed.

Second, a state of compulsive rumination need not be accompanied by the state of anti-hedonia that accompanies depression. It is not simply the absence of feeling good, it is active feelings of feeling bad, feeling so bad that death via suicide is perceived to be a desirable and sought for outcome. The only way that a state that induces a significant risk of death (as in suicide from depression) could evolve is if that state saves more lives that it costs. By what mechanism specifically does anti-hedonia decrease death sufficiently that deaths from suicide could be balanced and produce a net benefit?

Maybe individuals could be saved from death in trial by combat by giving up so they are spared by the victor. How does this translate into increased reproductive opportunities? If the fights are over females (essentially the only thing worth fighting to the death over), then depression should be sex-linked with females not exhibiting it. This is not observed. It should only be observed in sexually mature individuals, and not in children. It should not be observed in post-menopausal women, it should not be observed in the elderly. It should not be observed in individuals not undergoing mate selection stress.

I have a much simpler explanation. When an organism is being chased by a predator, the “fight or flight” state is triggered. In the extreme case, the organism can run itself to death. From an evolutionary standpoint, being caught by a predator and running oneself to death are “the same”, death and loss of all future reproductive possibilities. However the ability to run oneself to death is a very strong survival feature because the organism doesn’t need to run itself to death, it only needs to run until it has escaped from the predator. If the predator gets tired and stops before the organism runs itself to death, then the organism has survived and may have future opportunities to reproduce. Any injury or damage short of death is infinitely better than being caught by the predator. Death from exhaustion is only equivalent to being caught.

In order to induce a state where an organism can run itself to death, physiology must induce a state of euphoria. A state where the euphoria over rides all of the normal pain and fatigue warning signals so they can be ignored and the organism can willingly and euphorically run itself to death. I think this is the euphoria of the “runner’s high”, the euphoria of autoerotic asphyxiation, the euphoria of solvent huffing, the euphoria of stimulant abuse, the euphoria that near drowning produces. I call this the Euphoric Near Death State, ENDS. I see this state as brought about solely by extreme metabolic stress, near death metabolic stress.

During this physiological state, a state which may occur only a few times (or never) in an organism’s life span, there is not an emergency energy source that comes online, saved only for that emergency. That would be grossly inefficient. No, what happens is that essential systems get turned off for a short period of time. Things that are not needed while running from a predator can be turned off, and their metabolic resources diverted to running. It is the turning off of essential pathways that leads to eventual death from exhaustion. For example, is continued healing useful while running from a predator? No, if the organism is caught, that healing is useless. If the healing is turned off and the predator is escaped from, then healing can be turned back on. Any inefficiencies due to turning stuff off during the attempted escape are worth it if it produces escape. Any damage due to turning stuff off during the attempted escape is worth it if it leads to escape. It is this turning off of essential physiological pathways that causes death from exhaustion. There are degrees of course, and physiology is pretty good at ranking physiological pathways and needs so that only the longest time scale pathways are turned off first.

With healing and repair turned off, damage occurs and accumulated during this period. When enough damage accumulates, the organism dies.

Because the organism has to run while enduring extreme damage, broken bones, cuts, bleeding, muscle necrosis, etc, the state of euphoria has to be pretty extreme to counter all the signals that in other situations would produce inaction. Damaging a broken foot by running on it is a bad idea, unless a bear is chasing you.

Because the near death metabolic state has to be euphoric, and has to shut off all the “safeties” that normally protect an organism from self-injury, it is an extremely dangerous state for an organism to enter. If organisms could enter that euphoric state, at will, they would because it is an euphoric state. If they did enter that state at will, they would uselessly risk death. Organisms don’t enter that euphoric state at will, readily or easily. Therefore there must be an aversive state between the state “at rest” and the euphoric near death state. This aversive state provides a barrier to the easy entering of the euphoric state. It is my hypothesis that this necessary aversive state is the depressive state.

The hypothesis that depression is the aversive state between at rest and a near death metabolic state fits all the data. Depression is associated with reduced metabolic activity in the brain, reduced blood flow, reduced glucose oxidation, reduced ATP levels. It explains how metabolic compromise by itself (as in vascular depression) can cause depression. It explains the reduced metabolic activity, reduced immune system activity, lethargy, and other metabolic disturbances. It also explains the insomnia that sometimes accompanies depression. A state of stress, metabolic or otherwise is not a good time to sleep, if there are other things that can be done to reduce stress.

This hypothesis also explains all the disorders that are associated with depression, heart disease for example. If there is starting to be metabolic stress sufficient to induce depression, there aren’t enough metabolic resources to keep the heart and vasculature in good repair, so it degenerates.

The problem is that metabolic state that can produce depression can occur via multiple paths and that some states that are adaptive in the short term can cause depression in the long term. It is mostly a problem of the regulation of the metabolic state. The simplest way to look at it is as a problem of ATP regulation. ATP is exquisitely well regulated. It is so well regulated that a myth has developed about its regulation, the myth of homeostasis. I have a blog on that, from a few years back. Because physiology must control the ATP consumption of so many different pathways, and regulate them all “in sync” with ATP supply and demand, physiology must use the ATP level as a control parameter to regulate ATP consumption. At high ATP levels non-voluntary pathways like healing are fully on. At low ATP levels they are off, but voluntary pathways like muscle activity are on. This regulation is what allows muscle to consume ATP until it necroses. This is a feature in the short term, but to regulate the ATP levels in an entire tissue compartment (necessary to let all the cells work “in sync”), there must be a diffusible signal that communicates ATP status between adjacent cells. If this signal is not working properly, then the tissue compartment may enter a state of low ATP and be unable to leave it. It is the long term dis-regulation of ATP that causes depression.

Fix the ATP levels and you fix the depression. This is more difficult than it sounds. The primary stress response is to lower NO levels. This lowers ATP levels (via sGC) and raises the aerobic ATP generation rate by mitochondria (by disinhibiting cytochrome c oxidase). No doubt it also triggers many other stress responses. Ending the “fight or flight” state requires increasing NO levels. This is what is not easy. NO is extremely well regulated. There is no generally recognized way to raise NO levels. I think that topical ammonia oxidizing bacteria is the only way that will be effective.

In summary, depression is the feature that allows the near death euphoria that allows an organism to run itself to death while trying to escape from a predator. It is the feature that allowed the physiology to evolve that lets a parent to lift a car off their child. It allows all metabolic resources to be devoted to a single task for a short period of time, even if that diversion would cause death if extended for too long.

1. Depression is really code for some maladaptive or pathological condition rather than the whole class of persistent sadness which occur after significant loss (small-d depression in a sense). The study and alleviation of this big-D Depression is of great importance.

2. Most of the studies are of big-D Depression and probably have a disproportionately small cohort of small-d depression sufferers.

3. Big-D Depression is a condition, but small-d depression is more of a symptom.

In this way, we are talking about somewhat different but overlapping issues.

4. The idea that trauma and depression (either big- or small-d) are exclusive is bunk

5. I agree that my reference to Nolan-Hoeksema was a distortion of the 2008 article and her general concept of depression and rumination. But she did acknowledge both Watkins 2008 and also the possible value of rumination:

In light of the effectiveness of distraction and behavioral activation interventions for rumination and depression, it seems paradoxical that interventions designed to focus attention on distressing emotions and thoughts, such as experiential or mindfulness therapies, also have positive effects on depression in some studies…

6. My general POV is embodied-cognition, yet most of the discussion is in a more classical POV, so I resort to indulging my cognitive biases . Your writing and responses are much more deserving of the quality-impartiality labels.

Thanks for the comment! A couple of thoughts:

I believe that ‘biologically fit’ is a more parsimonious assumption than superfluous. It is quiet possible that however long humans have had awareness of distress, we have ruminated.

I suppose what “seems” adaptive to one is ultimately somewhat subjective (which is a major limitation of all of these just-so stories). My point was that if you posit that depression is adaptive specifically because it promotes rumination, you have a discontinuous hypothesis that doesn’t explain depression-like traits in other animals. It seems more plausible to me to have just one story that explains the range of depression-like behavior across the board, and then view the human-specific rumination as a by-product of another trait that clearly seems adaptive (intelligence). But I don’t really see any way to reconcile differing opinions on this point; unfortunately it’s not a question that lends itself to easy empirical validation.

I think it’s something of a mischaracterization to say that “Nolan-Hoeksama (2008) also accepts that rumination can have positive effects on depression”. Nolen-Hoeksema’s more recent work (and that of others) suggests that the effects of rumination aren’t uniformly negative (mainly because in some cases they’re neutral–not because they’re positive!), and that rumination bears some relation to more adaptive forms of reflection. But she specifically draws a distinction between rumination and adaptive self-reflection (which surely makes sense, unless you think that non-depressed people never have to solve self-relevant problems!). And her overarching point still remains that depressed people’s ruminative pattern is generally quite maladaptive; she emphasizes repeatedly (both in the 2008 paper and others) that there are all sorts of other cognitive distortions associated with ruminative tendencies. So on balance, the preponderance of evidence going back several decades (much of which she reviews) still clearly indicates that rumination is overwhelmingly a negative thing.

Trauma is often regarded as a violation of basic assumptions and believe. So recovery from trauma would, in part, involve the abandonment of old beliefs and the creation of new ones. The emotion sadness helps break down attachments and identifications, so in that sense, depression would have value. Constructive rumination, on the other hand, would be part of building new beliefs.

Recovery from trauma isn’t the same thing as depression. Actually, in many cases, it’s an exclusion criterion. If someone close to you dies suddenly, you may experience trauma and be very upset for a couple of weeks (that would be within the normal range of human response), but you generally wouldn’t consider that depression unless the person didn’t recover in a timely manner. So I’m not sure that’s relevant here. There’s also the question I alluded to with my cigarettes analogy of whether depression is more likely to lead to trauma in the first place (which is generally the case: dispositionally negatively people react more aversively to all kinds of major life events). So even if it were true that depression-induced rumination ultimately helps you get out of your bind (which, to reemphasize, I don’t think is the case), that’s not much consolation if your depressive tendencies are what got you into that bind to begin with.

Lastly, no one’s going to dispute that sadness often serves a valuable purpose, but there’s a world of difference between feeling sad (which everyone does from time to time) and being in a state of clinical depression. The Andrews and Thomson argument isn’t that sadness is adaptive, it’s that full-blown depression is adaptive specifically in virtue of promoting rumination. There’s very little evidence for the latter claim, and much against it.

At any rate, I appreciate the comment, and of course you can (and should) be skeptical of my skepticism.

it seems more parsimonious to see depressive rumination as a non-adaptive by-product of a more general and (potentially) adaptive disposition to experience negative affect.

I believe that ‘biologically fit’ is a more parsimonious assumption than superfluous. It is quiet possible that however long humans have had awareness of distress, we have ruminated.

Depression itself is positively correlated with posttraumatic growth (Dolbier et al. 2009). Watkins (2008) suggests rumination predicts reduced levels of depression. Nolan-Hoeksama (2008) also accepts that rumination can have positive effects on depression. The literature about rumination has shifted a fair amount in the past two years.

You write:

the very purpose of rumination might be to keep depressed people in a depressed state

Struggling with the consequences of trauma can be considered the main source of posttraumatic growth (Tedeshi & Calhoun 2004). Trauma is often regarded as a violation of basic assumptions and believe. So recovery from trauma would, in part, involve the abandonment of old beliefs and the creation of new ones. The emotion sadness helps break down attachments and identifications, so in that sense, depression would have value. Constructive rumination, on the other hand, would be part of building new beliefs.

Rumination is the consequence of trauma and distress. If you ruminate about “your co-workers hate you,” you are rumination of the very primal concern about the availability and sustainability of social-bonds and meaningful attachments. A trivial representation does not mean the source of the distress is trivial.

You site the losses which occur from depression. Those losses occur from trauma, and depression is one of its consequences. Depression itself can be a condition which supports recovery from the debilitating distress of the trauma.

One point, CBT is shifting in favor of a “mindfulness” component, and rumination is no longer regarded as something to be “broken out of” or suppressed. The idea is to be aware but not reactive. Trauma literature often discusses the concept of “window of tolerance,” and this CBT+mindful approach would have the quality of trying to help expand the window of tolerance and maintain the ruminations within it.

Can I be skeptical of your skepticism?