A study on depression

Con­tent note: de­pres­sion, suicide.

A friend I was dis­cussing de­pres­sion with linked me an ar­ti­cle, “Most an­guish isn’t an ill­ness but an evolved re­sponse to anx­iety”. It turns out it’s a con­densed ver­sion of a re­cent pa­per by Syme and Ha­gen, two biolog­i­cal an­thro­pol­o­gists. The ba­sic idea is that de­pres­sion, anx­iety and PTSD are not re­ally di­s­or­ders, but evolu­tion­ary adap­ta­tions, so we shouldn’t try to cure them (though we could try to re­move their causes).

I have analysed the ar­ti­cle and pa­per at length be­cause I wanted to give my friend a de­tailed crit­i­cism, and be­cause it ties in nicely with my hav­ing re­cently up­dated my be­liefs about psy­chi­a­try and start­ing to ques­tion some in­her­ited anti-psy­chi­a­try memes.

More­over, there is a whole class of ar­ti­cles and move­ments out there who say that “[in­sert men­tal con­di­tion] is not an ill­ness and we should stop med­i­cal­iz­ing it already”. To the ex­tent that they use similar ar­gu­ments, my anal­y­sis should hope­fully gen­er­al­ize.

De­pres­sion is not real. Is it?

The main prob­lem I have with this work is that the au­thors are si­mul­ta­neously deny­ing the con­cept of de­pres­sion and ap­ply­ing it in their anal­y­sis.

The ar­ti­cle starts by say­ing that view­ing de­pres­sion and anx­iety as ill­nesses is a “colos­sal er­ror”. In­stead, they are adap­ta­tions de­signed by nat­u­ral se­lec­tion and ac­ti­vated by en­vi­ron­men­tal fac­tors, which the au­thors mainly iden­tify with so­cial con­flict. They want to offer ar­gu­ments against the “dis­ease model” of de­pres­sion and anx­iety.

In the pa­per, they de­scribe some well-known limi­ta­tions of psy­chi­a­try. They at­tack the chem­i­cal im­bal­ance the­ory of de­pres­sion; point out that treat­ing the symp­toms is not the same as re­mov­ing the causes; that com­mon an­tide­pres­sants have limited effi­cacy, though they are more effec­tive than placebo; and that psy­chophar­ma­ceu­ti­cals have aver­sive side effects.

In Sec­tion 5, the au­thors com­plain of the DSM’s “stran­gle­hold on re­search and policy”. Un­like sci­en­tific tax­onomies that cut re­al­ity at the joints, such as the pe­ri­odic table of el­e­ments, the DSM’s clas­sifi­ca­tion sys­tem is based on clusters of symp­toms and is ba­si­cally a mess. The DSM has too many di­ag­noses, is vuln­er­a­ble to mar­kets and fads, shows low val­idity and tends to over-di­ag­nose in healthy pop­u­la­tions.

I think most of these crit­i­cisms are jus­tified, but the au­thors don’t provide any work­able al­ter­na­tive or new paradigm. In fact, af­ter com­plain­ing of their dom­i­nance, they keep us­ing the DSM cat­e­gories in their work as if noth­ing hap­pened.

Here we very ten­ta­tively offer a pro­vi­sional evolu­tion­ary schema with the im­por­tant caveat that be­cause most re­search has used DSM cat­e­gories, our schema re­lies on them too.

At the end of the pa­per, in their ad­dress to biolog­i­cal an­thro­pol­o­gists, they si­mul­ta­neously recom­mend and dis­cour­age the use of the DSM:

1. Con­sider in­ves­ti­gat­ing men­tal health is­sues in your study pop­u­la­tions. As a start­ing point there are sev­eral sur­vey in­stru­ments, based on the DSM or ICD, that have been de­vel­oped for cross-na­tional re­search.
2. Do not shackle your re­search to DSM/​ICD cat­e­gories and its symp­tom-based ap­proach. The DSM/​ICD symp­tom-based cat­e­gories were largely de­vel­oped based on in­pa­tient pop­u­la­tions in ur­ban psy­chi­a­tric hos­pi­tals in 19th and 20th cen­tury Europe and the United States.

What are their col­leagues to make of this?

“De­pres­sion” it­self is a psy­chi­a­tric cat­e­gory. If you wish to ar­gue that the con­cept has no val­idity, you should dis­solve it into other phe­nom­ena as per the re­duc­tion­ist ap­proach. The au­thors do at­tempt a re­duc­tion to biolog­i­cal adap­ta­tions, but their ap­proach is half-baked.

First, they don’t ex­plain why the illu­sion that there is such a thing as “de­pres­sion” arose in the first place. Why do states spend mil­lions so peo­ple can study – and psy­chi­a­trists learn to treat – a con­di­tion that doesn’t ex­ist that so peo­ple who don’t have the con­di­tion can get a cure they don’t need?

Se­cond, the au­thors keep us­ing the con­cept of de­pres­sion as a psy­chi­a­tric cat­e­gory. From the ar­ti­cle:

Globally, de­pres­sion is the largest con­trib­u­tor to the bur­den of men­tal ‘ill­ness’ by far.

A se­ri­ous feat for a di­s­or­der that is not a di­s­or­der. From the pa­per:

We there­fore pro­pose that there is a group of men­tal di­s­or­ders that are prob­a­bly best ex­plained by ge­netic-based de­vel­op­men­tal dys­func­tions; a group that is prob­a­bly caused by senes­cence; a group that might be caused by mis­matches be­tween mod­ern and an­ces­tral en­vi­ron­ments; and a group that are prob­a­bly not di­s­or­ders but in­stead are aver­sive and so­cially un­de­sir­able but nev­er­the­less adap­tive re­sponses to ad­ver­sity.

How can a “group of men­tal di­s­or­ders” con­tain things that are “prob­a­bly not di­s­or­ders”?

Put­ting com­mas around ‘ill­ness’ does not change the fact that you are an­a­lyz­ing de­pres­sion as an ill­ness. Deny­ing that de­pres­sion is a di­s­or­der does not change the fact that you are plac­ing it in a tax­on­omy of men­tal di­s­or­ders. Com­plain­ing against the “stran­gle­hold of DSM” is ab­surd if you keep us­ing its cat­e­gories un­crit­i­cally.

Fi­nally, a list of psy­chi­a­try’s prob­lems does not au­to­mat­i­cally provide an ar­gu­ment for its dis­man­tle­ment, es­pe­cially if you are not sug­gest­ing any work­able ideas for im­prove­ments or new paradigms.

De­pres­sion as adap­tive response

The au­thors be­lieve that the failures of psy­chi­a­try are partly due to its hav­ing failed to in­te­grate evolu­tion­ary the­ory. They pro­pose to use Wakefield’s con­cept of ill­ness as harm­ful dys­func­tion. Dys­func­tion in­di­cates the “failure of a trait to perform its evolved func­tion”. Harm­ful “sig­nifies an in­di­vi­d­ual, so­cial or cul­tural value judg­ment”.

Allegedly, this con­cept can “help dis­en­tan­gle true dis­ease states from con­di­tions, like ho­mo­sex­u­al­ity, that are merely con­sid­ered so­cially un­de­sir­able.”

Wait, didn’t they just say that “harm­ful” is an in­di­vi­d­ual, so­cial or cul­tural value judg­ment? How does that save ho­mo­sex­u­al­ity from be­ing con­sid­ered ‘harm­ful’ in the wrong con­text? It is ex­actly through an in­di­vi­d­ual, so­cial or cul­tural value judg­ment that ho­mo­sex­u­al­ity may be con­demned. To­day’s dom­i­nant judg­ment is that it’s not harm­ful, be­cause our val­ues have changed. But how would the harm­ful dys­func­tion ill­ness con­cept have pre­vented our fore­fathers from see­ing ho­mo­sex­u­al­ity as an ill­ness?

Their ap­proach to evolu­tion­ary the­ory is also prob­le­matic. Take the fol­low­ing quote:

It is an evolu­tion­ary para­dox that many men­tal di­s­or­ders are both harm­ful and her­i­ta­ble, be­cause dele­te­ri­ous alle­les should be elimi­nated by purify­ing se­lec­tion.

But I see no para­dox at all. First, the fact that some genes, pre­sent cer­tain con­di­tions, make peo­ple slightly more ex­posed to de­pres­sion in our twenty-first cen­tury so­ciety, does not im­ply that they had the same effect in the an­ces­tral en­vi­ron­ment. Imag­ine a gene that makes you de­pressed if you don’t hunt and kill big an­i­mals. This gene would make you de­pressed to­day if, like most hu­mans, you live in a city, but would have been quite adap­tive in the an­ces­tral en­vi­ron­ment. Se­cond, dele­te­ri­ous to whom? To the in­di­vi­d­ual? But evolu­tion is not about in­di­vi­d­u­als, it’s about genes. Genes that make the in­di­vi­d­ual mis­er­able can be neu­tral or fa­vor­able from se­lec­tion’s point of view (see last sec­tion).

The au­thors be­lieve that de­pres­sion di­s­or­ders, anx­iety di­s­or­ders and PTSD form a sep­a­rate cluster from other psy­chi­a­tric di­s­or­ders such as schizophre­nia, OCD, and bipo­lar di­s­or­der. This is jus­tified on the fol­low­ing grounds: (1) they have rel­a­tively low her­i­ta­bil­ity, (2) they are more preva­lent, (3) they tend to ap­pear at any age, (4) they tend to ap­pear in re­sponse to ad­ver­si­ties. The sup­port­ing data comes from a study by the Brain Con­sor­tium.

This is an in­ter­est­ing ob­ser­va­tion, of which I was not aware. Go­ing through the Brain Con­sor­tium study, I find sup­port for (1) and (2), while (3) seems ac­tu­ally falsified (see s.m., Fig. S1F), but I haven’t seen the other stud­ies that they’re cit­ing. I’m not sure about (4): are de­pres­sion and anx­iety more trace­able to ex­oge­nous causes than schizophre­nia and bipo­lar? In­tu­itively, I would say yes; it’s eas­ier to imag­ine some­one get­ting de­pressed or trau­ma­tized over an aver­sive event, than sud­denly be­com­ing schizophrenic or bipo­lar.

In fact, I am will­ing to ac­cept the au­thors’ claim that de­pres­sion ap­pears in re­sponse to ad­ver­sity. For ex­am­ple, they quote a study show­ing that

In con­flict-af­fected coun­tries, an es­ti­mated one in five peo­ple suffers from de­pres­sion, PTSD, anx­iety di­s­or­ders, and other di­s­or­ders, com­pared to 1 in 14 wor­ld­wide (Char­l­son et al., 2019).

The prob­lem is in their con­clu­sion:

Be­cause many of their symp­toms seem to be func­tional re­sponses to threats, the hy­poth­e­sis that they are func­tional re­sponses to ad­ver­sity is com­pel­ling.

Note the tau­tologic­ity of this state­ment. Like­wise, in the ar­ti­cle:

Con­sis­tent with our pre­dic­tions, ev­i­dence from lon­gi­tu­di­nal stud­ies in the United States and Europe in­di­cates that ma­jor de­pres­sion, PTSD and anx­iety mostly fol­low ad­verse life effects, im­ply­ing that they are of­ten a re­ac­tion or defence against those events.

How does effi­cient cause au­to­mat­i­cally trans­late to func­tional cause? It’s a long stretch from “de­pres­sion fol­lows ad­verse life events” to “de­pres­sion is an adap­tive re­sponse to ad­ver­sity”. At the very least, it seems that the propo­si­tion (de­pres­sion arises be­cause of X AND de­pres­sion is a func­tional re­sponse to X) is more de­tailed than (de­pres­sion arises be­cause of X) and hence would need more proof.

The au­thors’ strat­egy is to equate de­pres­sion to pain and then show that pain is an adap­tive re­sponse.

per­sis­tent sad­ness, low mood and and­he­do­nia [...] are prob­a­bly forms of “psy­chic pain” that adap­tively fo­cus at­ten­tion on ad­verse events that would have re­duced fit­ness [...] so as to miti­gate the cur­rent ad­ver­sity and avoid fu­ture such ad­ver­si­ties.

This is very clear in the ar­ti­cle:

Just be­cause psy­cholog­i­cal pain is un­pleas­ant to the self and oth­ers, that doesn’t make it a dis­ease, and we shouldn’t seek in the first in­stance to blunt it with drugs or other med­i­cal in­ter­ven­tions.
[...] psy­cholog­i­cal pain, like phys­i­cal pain, prob­a­bly evolved by nat­u­ral se­lec­tion, and in many or most cases is there­fore not a dis­ease.
[...] evolu­tion has not shaped hu­mans to be per­pet­u­ally happy or free of pain.

(I must have missed the lat­est edi­tion of the DSM which clas­sifies de­pres­sion as not be­ing made of he­do­nium).

The au­thors take the easy way out by ig­nor­ing the spe­cific (and I would add, mal­adap­tive) fea­tures of de­pres­sion. In fact, any men­tal prob­lem can be de­scribed as adap­tive, be­cause it in­evitably in­volves some emo­tion or other, and emo­tions are adap­tive be­cause they were de­signed by evolu­tion, weren’t they? So anx­iety is char­ac­ter­ized by “func­tional re­sponses to gen­uine so­cial and en­vi­ron­men­tal threats”. PTSD shows “func­tional re­sponses to avoid fu­ture trau­mas”, al­though the au­thors ad­mit that other symp­toms “are not clearly func­tional” (but they might be!). Even post­par­tum de­pres­sion is “an adap­ta­tion that in­forms moth­ers that in­vest­ment in a new offspring is un­likely to de­liver fit­ness benefits”.

Any­one can come up with a just-so story of how a con­di­tion is adap­tive, but this does not con­sti­tute a proof. To prove that de­pres­sion (or any other con­di­tion) is an adap­tive re­sponse to averse events, one should at least: (1) show that there were spe­cific evolu­tion­ary pres­sures to de­velop the adap­ta­tion, (2) de­scribe the mechanisms that ex­e­cute the adap­ta­tion, (3) show which fac­tors in the cur­rent en­vi­ron­ment ac­ti­vate the adap­ta­tion.

The sub­ject abandoned

The De­pressed Per­son goes to the psy­chi­a­trist for her ap­point­ment.

Out come Syme and Ha­gen:

“You’ve noth­ing to do here. Go home.”

“But… I was sup­posed to see Dr. Smith.”

“There is no rea­son for you to see him. He’ll just stig­ma­tize you and mess with your brain. Be­sides, you’re not even ill.”

“But… I feel ter­rible. Most days I can’t even get out of bed. Dr. Smith said I have ma­jor de­pres­sive di­s­or­der.”

“And you be­lieved him? Haha. Listen, don’t beat your­self up. There’s noth­ing wrong with how you’re feel­ing. It’s an evolu­tion­ary adap­ta­tion, you see?”

“But… It feels wrong.”

“Of course it does! We didn’t evolve to be in a per­pet­ual state of or­gasm, did we? Tell me now, did some­thing par­tic­u­larly hor­rible hap­pen to you re­cently?”

“Well, yes. I was as­saulted and mugged on the street.”

“There! A perfectly adap­tive re­sponse to so­cial con­flict! Now, what would hap­pen if you didn’t feel any pain at all ?”

“Um… I’d feel a lot bet­ter?”

“Non­sense, non­sense. Listen, we’d love to stay and chat, but there’s a psy­chi­a­trist who’s try­ing to com­mit a suici­dal per­son on the other side of town.”

Crit­ics of psy­chi­a­try typ­i­cally say that it de­val­ues peo­ple by stig­ma­tiz­ing them, la­bel­ing them as dis­eased, mess­ing with their brains, clos­ing them in in­sti­tu­tions and so on. They, on the other hand, tend to style them­selves as the defen­ders of the dig­nity and au­ton­omy of the sub­ject.

Con­sider what hap­pens to the de­pressed per­son in Syme and Ha­gen’s world. Their con­di­tion can­not be ad­dressed be­cause it is not an ill­ness (no in­surance for you, sorry) and there is no point in try­ing to cure an adap­ta­tion. Hence, the in­di­vi­d­ual is left to their own de­vices. If they ask what they should do to feel bet­ter, one can do lit­tle more than shrug: “Noth­ing much. Not un­til we man­age to erad­i­cate the roots of so­cial con­flict, at least.”

If you’re one of the 7% of peo­ple who are de­pressed, you should hope that they are wrong and you don’t live in that world, be­cause if you do, your situ­a­tion is quite hope­less:

If our hy­poth­e­sis that de­pres­sion is an adap­tive, evolved defence is true, it wouldn’t offer much hope for a ‘quick fix’ or med­i­cal cure. [...] Just be­cause psy­cholog­i­cal pain is un­pleas­ant to the self and oth­ers, that doesn’t make it a dis­ease, and we shouldn’t seek in the first in­stance to blunt it with drugs or other med­i­cal in­ter­ven­tions. In­stead, we should look to the so­cial roots of ad­ver­sity – to in­equities, in­jus­tices and in­di­vi­d­ual self­ish­ness – and con­sider if and how we can har­ness men­tal an­guish to help change our­selves, and other peo­ple’s lives, for the bet­ter.

Oh, and not only is your con­di­tion not an ill­ness – you should har­ness your men­tal an­guish to im­prove your­self and other peo­ple’s lives. What are you achiev­ing with your de­pres­sion?

Pan­glos­sian evolutionism

The goal of theod­icy is to con­nect evil and re­pul­sive phe­nom­ena to the benev­olent in­fluence of God. No mat­ter how hor­rible and ab­surd some­thing is, it was made by God, and God has de­signed ev­ery­thing, and God is good, so that thing must be good too; you’re just not see­ing its good­ness.

Analo­gously, there is a sort of pan­glos­sian evolu­tion­ism that aims to show that some un­de­sir­able con­di­tions are ac­tu­ally good be­cause they were put there by evolu­tion, and evolu­tion has de­signed ev­ery­thing, and evolu­tion is good be­cause it wants us to sur­vive, so the con­di­tion must be good too; you’re just not see­ing its func­tion.

The 1755 Lis­bon Earthquake may be the most in­fluen­tial earth­quake in the his­tory of civ­i­liza­tion. It en­couraged the first em­bry­onic stud­ies of seis­mol­ogy. It changed the course of Por­tu­gal’s his­tory. Most im­por­tantly, it in­spired Voltaire to ridicule Leib­niz’s theod­icy, lay­ing a cor­ner­stone of en­light­en­ment think­ing. The earth­quake struck on All Saints’ Day, kil­led tens of thou­sands, and de­stroyed most of the city’s churches. What benev­olent God would do this?

The moral of the story is that pan­glos­si­anism will not sur­vive a close en­counter with Na­ture.

From The Selfish Gene (chap­ter 7):

Some­times, as we have seen, one mem­ber of a lit­ter is a runt, much smaller than the rest. He is un­able to fight for food as strongly as the rest, and runts of­ten die. We have con­sid­ered the con­di­tions un­der which it would ac­tu­ally pay a mother to let a runt die. We might sup­pose in­tu­itively that the runt him­self should go on strug­gling to the last, but the the­ory does not nec­es­sar­ily pre­dict this. As soon as a runt be­comes so small and weak that his ex­pec­ta­tion of life is re­duced to the point where benefit to him due to parental in­vest­ment is less than half the benefit that the same in­vest­ment could po­ten­tially con­fer on the other ba­bies, the runt should die grace­fully and will­ingly. He can benefit his genes most by do­ing so. That is to say, a gene that gives the in­struc­tion ‘Body, if you are very much smaller than your lit­ter-mates, give up the strug­gle and die’, could be suc­cess­ful in the gene pool, be­cause it has a 50 per cent chance of be­ing in the body of each brother and sister saved, and its chances of sur­viv­ing in the body of the runt are very small any­way. There should be a point of no re­turn in the ca­reer of a runt. Be­fore he reaches this point he should go on strug­gling. As soon as he reaches it he should give up and prefer­ably let him­self be eaten by his lit­ter-mates or his par­ents.

As I read this pas­sage, my im­me­di­ate thought was: Maybe this is why we have de­pres­sion. My sec­ond thought was: It’s so hor­rible it just might be true.

If there was an evolu­tion­ary ex­pla­na­tion for de­pres­sion, this is ex­actly what I’d ex­pect it to sound like. Perfectly log­i­cal and perfectly hor­rible.

Re­mem­ber, se­lec­tion is about genes. The in­di­vi­d­ual is an adap­ta­tion ex­ecu­tor, a ma­chine meant to be dis­carded when it has ex­hausted its pur­pose. If you can’t grow and re­pro­duce, you might as well try to benefit the copies of your genes that will live on in your kin.

This is what de­pres­sion would sound like in Dar­wi­nian idiom: “Body, you are use­less. Lose all hope and in­ter­est in life. Stop eat­ing and drink­ing. Lay down and die already so that your mother and siblings can eat you.”

Turns out the idea of de­pres­sion-as-adap­ta­tion is not that com­fort­ing af­ter all.