03 October 2014

Evolution, Depression and Suicide.

Is it possible that "mental illness" is an evolutionary adaptation to prevent us committing suicide when we feel like ending it all? It seems unlikely, but this is the subject of an article tweeted by @sarahdoingthing September 28, 2014.
C A Soper. "Anti-suicide mechanisms as a general evolutionary explanation for common mental disorders". University of Gloucester. [The article appears to be formatted for a publication, but no publication details are included. It's hosted at academia.edu where you can also find my academic work]. 
Since I have a long interest all three subjects (evolution, depression and suicide) I started to try to discuss the article on Twitter, but could not compress my thoughts into 140 characters. So this essay critically examines Soper's approach to evolution. 

Soper argues that (1) "Common mental disorders are too common not to exist for a reason." and that (2) "On this evidence it is reasonable to deduce, as evolutionary psychologists do, that common psychiatric disorders must have their origins in natural selection" and that (3) "When two traits routinely occur together in this way, it is reasonable to infer that a single process is at work behind both, and a mechanism offered to explain one must be flawed if it cannot also explain the other." I want to look at these three underlying claims specifically and consider whether this is credible approach to evolutionary psychology and thus whether the conclusions drawn are valid.

Evolutionary Psychology

The field of evolutionary psychology is the approach of trying to find evolutionary explanations for (i.e. applying the idea of natural selection to) the ways our minds work. If our minds can be shown to work in a particular way, then we assume that this has some evolutionary advantage that has been selected for in the sense that people with this trait are more likely to survive and produce viable offspring. Evolutionary psychologists try to identify what that advantage is or speculate on what it might be. The second prong to this approach is to try to locate genes specifically associated with this function. This has produced the, usually spurious, media syndrome of reporting that "the gene for X has been found". Soper is not concerned with the second approach, only with trying to explain a puzzling phenomenon of the co-morbidity of some well-known problems.

On the whole I find evolutionary psychology a compelling hermeneutic. The idea that traits emerge and are preserved not just in our anatomy, but also in our neuro-anatomy and therefore in our brain function and behaviour, and that this anatomy is determined by our genome, all sounds quite reasonable. Amongst the more credible proponents of this theory include Robin Dunbar, Justin Barrett, and Robert McCauley. Dunbar has cogently argued that behaviours like laughter, singing and dancing helped to lower the time burden of maintaining social relationships (by enabling one to many relationships that replaced one to one grooming in ways not available to our ancestors or to present-day chimps) and made living in larger groups practical after our increased neocortex size made it possible. Larger groups have benefits in terms of protection from predators and thus help individuals in groups to survive. Justin Barrett has argued that our predilection for seeing agency in events makes us alert to being hunted and allows us to avoid becoming food for a predator, but as a side effect also primes us for believing in supernatural agency. And so on. 

Soper's main sources for the evolutionary approach seem to be textbooks specifically related to evolutionary psychiatry. I'm not familiar with these authors or their work so I can't comment on them, though I am surprised not to find reference to more fundamental research on evolutionary psychology in his article.

Mental Disorders and Evolution

Soper's specific claims begin with this: "Common mental disorders are too common not to exist for a reason." By "reason", here, he means for a positive reason. Soper wants to argue that every common trait must, by definition, give us an evolutionary advantage or it wouldn't have survived (which is the most simplistic reading of evolution). But in this he is wrong. For example hair colour and eye colour are common traits and there's no plausible evolutionary advantage to having different coloured hair and eyes. By way of contrast, we know that skin tone is directly related to long term habitation in certain latitudes. If a group lives on the equator for a few thousand years their skin will be darkened by melanin. And if a group lives at 50 degrees north for thousands of years their melanin decreases and they become pale. It's a positive adaptation to the amount of sunlight and vitamin D synthesis and it occurs over relatively short time scales as evolutions goes (and thus is probably epigenetic - a change in gene expression rather than a mutation of a gene). Importantly it makes a mockery of the concept of colour-based race. One must always be alert to other sources of change or variety.  One thinks also of the impact of our microbiome (the sum total of microscopic life that lives in and on our bodies and plays significant and often vital roles).

Think also of the common trait of susceptibility to being infected by viruses. On Soper's logic -- that susceptibility to depression conveys an evolutionary advantage -- our susceptibility to viral infections, such as influenza or ebola, is so common that it must also confer some evolutionary advantage. Viruses exploit vulnerabilities in surface features of our cells that evolved for other reasons. For example sperm use a similar mechanism to deliver their DNA to an ovum. These diseases are virulent and indiscriminate. Before modern medicine an influenza outbreak could kill millions of people. The 1918 influenza pandemic which killed 50-100 million people worldwide is a good example. And influenza is constantly mutating so that there is no immunity conferred from having had the disease once. The best evolutionary argument might be that such diseases weed out the weaker members of the species: i.e. that influenza is natural selection in action. This has the indirect advantage of allowing stronger members to live with less competition. In the simple version of evolution, then, we have positively evolved to allow weak members of our species to be eliminated by disease, though I don't find this a compelling argument and this the opposite of what Soper is arguing for depression. Unlike viral disease which reduces competition by killing weaker members of the species, Soper is arguing that mental illness, specifically depression, prevents those who are susceptible to suicidal ideation from actually committing suicide. Thus it acts to preserve people who carry a trait that in the cold light of day makes them less fit in an evolutionary sense (and I say this as a life-long sufferer of depression). On the face of it Soper is describing anti-evolution.

As already suggested Soper makes an extraordinary assumption in his view of evolution. He assumes that evolution is the only force at work on our mental states. Another aspect of our history he takes no account of is the massive changes that began to occur ca 12000 years ago as our ancestors began to form stable settlements: i.e. civilisation. As Robin Dunbar points out (Human Evolution) for primates being in large groups of strangers is stressful. Of course we find ways of coping with that stress - clear evidence of alcohol use begins around the same time as large scale settlements in Anatolia. 12,000 years is enough time for our skin to change the levels of melanin produced, but it is not enough time for major changes to the genome, especially under the kind of NeoDarwinian paradigm that Soper unquestioningly adopts. Dunbar, again, notes that we are evolved to live in groups of ca 150 with progressively weaker links to units of ca. 500 and ca. 1500. just as present day hunter-gathers still live. Limits are imposed on group size by the amount of neocortex in the brain which has not changed significantly in the last 200,000 years since anatomically modern humans first emerged in East Africa. We cannot keep track of more than 150 relationships (on average) and groups considerably larger lose coherence. Indeed in present day hunter gatherer society most groups spend the night in groups of about 50 that have close links (often by marriage) to two other groups of 50. City dwellers are forced to adapt to their situation by adjusting how much time they spend on the different layers of their social structure (generally more time spent with less people), but the average number of Facebook "friends" is still ca. 150.

Thus simply living in settlements creates enormous stresses on humans that no other primate has ever faced. Since civilisation brings many changes in terms of how we spend our time (eps. work) and what we eat (esp. the gross over-availability of calorie rich foods) it is clearly one of the most important factors in considering the health, mental or otherwise of modern humans. In many ways one could argue that we are not well adapted to modern life - slumped over a keyboard developing bad posture, carpal-tunnel-syndrome and occupational overuse syndrome, while gorging on foods laden with fat, salt and sugar so that we overflow the poorly designed chairs we sit on for most of our sedentary day is hardly an advert for evolution. If anything many of us are not evolutionarily fit for this environment and increasing numbers are having civilisation-related or "life-style" illnesses like coronary heart disease, type II diabetes, etc. 

One of the strengths of Professor Robin Dunbar's work is his ability to compare his results with other primates and to extract evidence from fossilised remains. It allows him to take a genuinely evolutionary view of the traits he is examining by showing how things have changed over time. When we only examine modern humans, have no reliable data for change over a time scale beyond ca. 50 years, and have little reliable data from outside Europe and America the method is very much weaker. Soper presents no data from other primates on mental illness and suicide for example. I suspect that this is because there is none. Animals don't, on the whole, deliberately kill themselves though they do show analogues of some kinds of mental illness and are susceptible to addiction (at least in laboratories). 

The first challenge of any evolutionary study of suicide is to try to determine when humans began to kill themselves. And of course it's impossible to tell because the kind of evidence we need is unavailable. So the theory that we evolved this behaviour is already on very shaky ground. There is no history, no fossil record, none of the evidence over time that is crucial to all evolutionary arguments. The second challenge is to explain why humans do and other primates do not kill themselves. No explanation is presented for this either, except that Soper simply states that it must have an evolved in humans. Indeed he treats present rates of suicide as evidence of suicidality as part of "human nature". Now there is a slippery concept if ever there was one: human nature. It's entirely out of place in a scientific article. And the idea that present data represent historical data is simply mistaken. All we know for sure is that there are some ancient literary records of suicide (see my article Suicide as a Response to Suffering for a survey of suicide in the Pāli texts; where, coincidently, alcohol is described as leading to madness). We can associate suicide with settled human culture, for a few thousand years, but there is no evidence whatever for the evolution of suicide, it's simply an assumption that everything evolved because in the paradigm every trait is the positive result of evolution. 

"Other addictive, obsessive and compulsive behaviours may function as dis-tractions, effectively keep-ing a person in danger of suicide mentally and physically preoccupied.

Depression may be understood equally as a means to incapacitate a potentially suicidal indivi-dual:" (Soper p.2)
Soper cites a number of opinions on suicide and its aetiology, but noticeably absent is the monumental, if a little dated, study of suicide by Durkheim. One of Durkheim's main points is that suicide seems to be strongly associated with social isolation. This jibes well with other evolutionary psychology authors. As social animals we thrive if and only if we are part of a thriving community. Modern humans evolved for participation in a community of ca. 150 people. In fact we moderns frequently live in massive conglomerations of hundreds of thousands, if not millions, of people almost all of whom are strangers. Modern life allows a sizeable minority to become isolated and alienated from society. Many moderns live in social isolation to some degree. We're surrounded by strangers and have none of the intimate exchanges that bond primate groups, not even the sublimated activities of laughing, singing, dancing or praying together (cf Dunbar). What the effects of this have been over the long term, we are only just beginning to understand. Clearly some thrive in this new configuration, but some do not. And those who do not, I would argue, are those who develop so-called mental illness. Of course there are other, often organic, causes of mental illness as well and this is not a causal argument yet, but highlighting a correlation that begs to be investigated. Importantly, there is no unitary phenomenon here that can be ascribed to a single simple cause. Not even the depression that Soper focusses on has a singular aetiology. But Durkheim's original observations on suicide seem to stand up. 

Soper argues that (3) "When two traits routinely occur together in this way, it is reasonable to infer that a single process is at work behind both..." By two traits here, Soper is specifically referring to addiction and depression. His solution is argue that depression, with its associated lethargy, contributes to suppression of the suicidal ideation that occurs in the addict. This assumption appears to stem from his conclusion, not the other way around. He also flirts with the fallacy that correlation indicates causation. Certainly a strong correlation is interesting and deserves further study, but I doubt it is reasonable to infer from the outset that a common mechanism is at work when there's no common mechanism for depression in it's various forms nor one for suicide.

Importantly Soper presents a caricature of depression as involving lethargy. But he does not account for the phenomenon of depression associated with irritability and anger that is common, but under-reported and poorly understood, in men. Cf. Irritability, Anger Indicators of Complex, Severe Depression; or Depression & Men. Indeed the popular media representation of depression often focusses on women who have a big collapse, can't get out of bed for 6 months and then recover. That's not typical of depressed men, nor for the people who suffer repeated bouts of major depression or those who suffer long-term depression. The different aetiology of depression in men may be why men are twice as likely (15 per 100k) than women (8 per 100k) to commit suicide (WHO). Some will say that men need to talk about their problems more, but this is a simplistic and unhelpful generalisation. I've commented on this elsewhere so won't say more here. But if a supposedly singular problem is characterised by at least two unrelated traits (lethargy or anger), manifests differently in the sexes, and can be acute or chronic, then we've most likely been too superficial in our explanation and need to look more deeply. There's no one problem called "depression".

One of the most important and productive ways of looking at depression is to see the popular "chemical imbalance" explanation as having a behavioural cause. Over-stimulation of various brain mechanisms leads to problems. Constant anxiety—with activation of flight-or-fight response—can lead to lethargy and unresponsiveness, both characteristics of depression (I first experimented with this by examining the fight-or-flight response of earthworms more than 30 years ago for a high-school science class). Over-stimulation of pleasure mechanisms (through drugs, porn, eating, etc.) leads to an inability to experience pleasure—both through endorphin mediated pleasure/well-being, and through dopamine mediated anticipation and reward—also characteristic of depression. I can offer no explanation of the anger or rage felt by depressed men as yet.

One observable result is consistently lower serotonin levels in depressed people. But even after many decades there is no evidence for a causal relationship between serotonin (a hormone that has multiple roles in the body) and depression. Indeed the fact that antidepressants raise serotonin levels almost immediately, but (when they do work) take two to four weeks to lift mood, suggests something far more complex is going on.

Soper is also interested in the co-morbidity of depression and addiction. Robin Dunbar makes an interesting aside in Human Evolution. Alcoholics do not become addicted to alcohol per se, they become addicted to the endorphins that alcohol stimulates. Endorphins are one of the primary hormones produced in primates by mutual grooming and produce the sense of well being and contentment that comes from being a well established group member. Laughing, singing, and dancing in groups have the same hormonal effect. We're 30 times more likely to laugh at a comedy in a group than we are alone. This is consistent with the neuroanatomy of pleasure that I outlined in The Science of Pleasure, based largely on a book called The compass of Pleasure by David Linden (well reviewed here). See also my 2013 essay Pleasure, Desire and Buddhism

Addicts, according to David J Linden's recent account of addiction, overstimulate the part of their brain that is also responsible for the feelings of well-being associated with positive social interactions. Addicts who over-stimulate this function, progressively become unable to experience that feeling of well being, or only associate it with their drug of choice (the exception being nicotine addicts who use the frequent but weak stimulation of smoking as a way of bonding). There are in fact at least two mechanisms working in tandem: addicts gradually become less able to experience well-being and/or pleasure in the absence of their drug; and they make poor decisions and become unreliable as a result of the effects of the drug and thus become socially isolated. All too often drug abuse is initiated by some lingering unhappiness or dissatisfaction that might have led to, or already caused, depression anyway. The obvious example is that abuse and neglect in their various forms, especially at crucial developmental stages, can leave people vulnerable to depression.

By Robin Dunbar's argument, social alcohol use persists, despite the risk of addiction in some people, because it plays an important role in allowing us to operate in larger groups than we would otherwise have time for (with all the benefits that large groups provide). Disinhibition makes for fun, laughter, singing and other promoters of a sense of well-being and communality. This is not natural selection in the usual sense, in that we are not genetically programmed to make and consume ethanol, but it is natural selection in that societies which used alcohol to enhance social bonding seem to have prospered.

These mechanisms that mediate the experience of pleasure, well-being, and anticipation and reward clearly have evolved and we know them in quite a lot of detail now: which areas of the brain are involved, when those areas evolved, which neurotransmitters are involved and the more generalised impact of disrupting these mechanisms. Any evolutionary approach to mood disorders or addiction really needs to get to grips with these mechanisms and show how they are involved, preferably by citing clinical evidence, just as Dunbar and Linden do and Soper does not.

Soper speculates that addiction might "distract" the person from acting on suicidal impulses. Some addicts use substances in an attempt to control how they feel, to compensate for the lack of pleasure or reward or to suppress feelings of shame or anger. But is this really an evolutionary argument? Does our potential to abuse substances really convey an advantage? In the end the substance of choice in addiction is often the means of suicide (albeit slowly), just as many depressives over-dose on their anti-depressant medication. Is the alcoholic who does not commit suicide, but whose behaviour causes the breakdown of supportive familial and working relationships, and who suffers liver and brain damage really ahead on points? The deleterious effects of drugs during pregnancy are so severe (e.g. Fetal Alcohol Syndrome) that they must surely outweigh any perceived advantage from merely being alive to pass on one's genes. Soper's argument here is facile at best.  


Soper's key claim is that the lethargy commonly associated with depression acts as a defence against suicidal ideation. Maybe. But we also know that people with depression are far more likely to commit suicide that people without it. So if it is a mechanism, it's not a very good one. One suicide prevention website reckons that [in the USA] "15% of those who are clinically depressed die by suicide." and "The strongest risk factor for suicide is depression." (Save) Suicidal ideation and impulses are one of the most common features of the experience of depression.

Suicide is a terrible problem. It is the fifteenth most common cause of death worldwide (WHO). Depression is a leading cause of suicide (and to date I think it is under estimated because of the failure to fully recognise how it affects men). When someone kills themselves their family and friends are often left shocked, sad and angry. Suicide often seems like a betrayal. On top of everything, people are angry because the suicide has broken off the relationship, has not reached out, has not apparently reciprocated the love they feel. Death is never easy, but to most people suicide seems so preventable because it involves a conscious choice. As @sarahdoingthing says it's hard to understand because it is sui generis (self generated). How do the living empathise with the wish to be dead? Mostly they do not. The difficulty is to see how the choice is made by a disordered mind in a person who has frequently lost the ability to experience a sense of connection and does not have the perspective to see that the situation is temporary. Depression feels like solitary confinement.

Because it's so difficult to imagine what depression or addiction is like, most people who have not experienced it find they cannot empathise easily with sufferers. Very often the problems are ascribed to personal weakness such as a "weak will" or a moral failing (an example of the fundamental attribution fallacy). This can have the effect of increasing the social isolation of the person afflicted with depression. This is part of the stigma of mental illness.

The best thing you can offer someone who suffers is to listen to them without judgement and deal with your own discomfort discreetly. Whatever you do, don't offer unsolicited advice. If you're concerned about someone's safety encourage them to seek professional help. If you feel certain someone will harm themselves take whatever action you feel appropriate, but don't expect to be thanked (at least not right away).

In any case we need to be careful when constructing arguments based on evolution. It is no doubt a powerful and at present fashionable explanatory framework. There's no doubt in my mind that we evolved into our present form. But modern humans are unusual in the animal world in having the ability to over-ride evolution using culture and civilisation. While our genes are the blueprint for our neuro-anatomy, experience is a powerful shaper (both literally and figuratively) of the brain.

Without clear evidence of change over time evolution is a weak explanation. It may well be the explanation, but we cannot show why. Sometimes a trait has an obvious evolutionary advantage - language, mentalising, and laughter all provide demonstrable advantages and fit well with other areas of the theory. The potential to suppress suicidal impulses might confer an advantage, or it might have a deleterious effect on the population. Who is to say that suicide is not an instrument of natural selection? How do we weigh up the costs and benefits in these complex problems? I find no answer in Soper's article.

Even though we can identify commonalities depression and addiction likewise have multiple causes. When combining traits with many causes we multiply the complexity. Seeking unitary causes for complex problems is understandable, but often leads to fallacious thinking. Seeking a single generalised evolutionary explanation in terms of conferred advantage looks ideological. And in this case the premise looks flawed at best. So the interpretation of the data is unlikely to be trustworthy. For all these reasons I find Soper's theory unconvincing. Scarily, Soper is already making suggestions on implications for therapy as though his theory was sound. 



Barrett, J.L. (2004) Why Would Anyone Believe in God? Walnut Creek, CA: AltaMira Press.

Dunbar, R.I.M. (1992). 'Neocortex size as a constraint on group size in primates.' Journal of Human Evolution 22 (6): 469–493. doi:10.1016/0047-2484(92)90081-J 

Dunbar, Robin. (2014) Human Evolution: a Pelican Introduction. Pelican.

Durkheim, Emile. (1897) Suicide : a study in sociology. The Free Press, 1951.

Jayarava. (2004) 'Suicide as a Response to Suffering.' Western Buddhist Review.

Linden, David J. (2011) The Compass of Pleasure: How Our Brains Make Fatty Foods, Orgasm, Exercise, Marijuana, Generosity, Vodka, Learning, and Gambling Feel So Good. Viking.

C A Soper. "Anti-suicide mechanisms as a general evolutionary explanation for common mental disorders". University of Gloucester.

Related Posts with Thumbnails