Ed Hagen recently wrote a paper outlining his objections to the classification of major depression as a “brain disorder,” on the grounds that, in sum: the diagnosis is made to distinguish it from other conditions and not from “normal” persons, symptoms of what is called depression tend to remit within weeks or months, and occur at points in life where some amount of sorrow would be expected, and depression is a continuous distribution, so the cut-off for disordered behaviour is arbitrary. Although I cannot disagree with any of this, I think it misidentifies the problem, which is now potentially avoidable thanks to recent advances in genetics.
Namely: psychiatric diagnoses are based upon symptoms and not their genetic place of origin, evolution, or adaptive value.
By definition, natural selection only allows adaptive or neutral alleles to stay around; alleles can only avoid extinction if they keep up with the rate at which competing variants are reproduced. Thus, when an organism displays maladaptive behaviours, the explanation for this falls into one of three categories: deleterious mutations (the occasional fuck-up in DNA’s copying process), pathogens, and gene-environment mismatch. The last of these refers to situations in which genes that are adaptive in some environment are still present as the environment changes – they simply have not had time to be selected out yet.
As Hagen notes, the symptoms of depression are usually synonymous with the symptoms of neuroticism – a personality variable which remains fairly constant throughout the lifespan and determines responsiveness in aversive situations such as the death of a first-degree relative. However, it is not as though there is no variation in trait neuroticism itself; some people and groups are known to be higher in it than others, e.g. women. How much of that variation is “normal,” in adaptive terms? Perhaps very little, I suspect.
Some of it is obviously gene-environment mismatch. We have not had planes or parachutes for that long, which is why most people are more scared of skydiving than driving a car despite the fact that the latter is demonstrably more dangerous. Equally, sex differences are generally a sign of different (historical) adaptive challenges for the sexes, which may be why women are more neurotic than men by ~0.4 standard deviations, roughly equivalent to two subpopulations of males with a mean height difference of 1¼ inches – think of the English vs. the Spanish. Barely noticeable at the mean, but very much so at the tails. But, on the whole, I doubt that most variation in neuroticism is adaptive.
The genetic architecture of personality traits looks similar to intelligence in that both are massively polygenic and only a small chunk of the variance is eaten up by the “common” neutral alleles. In the case of personality, it might not even be as much as 10%. The rest, according to this paper, is due to “rare variant effects and/or a combination of dominance and epistasis.” These common (freq. > ~1%) variants are in a kind of equilibrium because each has reproductive costs and benefits, otherwise it would be impossible for them to all be common. For a personality trait such as agreeableness, it may be for example that genes which inculcate high agreeableness make one less attractive at the outset, especially as a male, but more fecund in the long run because agreeable people are more willing to have more kids, etc. The rest of the variance, in the individually rare (≤ 1%) alleles, will be deleterious, hence their rarity.
A deleterious allele can accumulate in the population until it reaches equilibrium frequency, the point at which further accumulation is counterbalanced by selection. The equilibrium frequency for a given allele is generally just the mutation rate at its locus divided by its reduction in fitness relative to the population average, e.g. if the population’s average birth rate is 2.0 and the allele knocks carriers down to 1.98, that is a fitness loss of 1%. For an allele with a mutation rate of 0.0001, this gives you an effective “maximum” frequency of 1%. Given the number of variants involved in the brain, there are apparently a lot of these, almost everyone is carrying some, and the unluckiest, at the right tail of mutational load, could be carrying quantities orders of magnitude more than the average.
Since the behavioural correlates of neuroticism are not neutral and tend heavily towards the maladaptive (references: 1, 2, 3), one has to question how much of “normal” sorrow, grief, and anxiety is really normal. Common, sure, but nonetheless aberrant. Natural evolution does not offer a straightforward means to eliminate it in toto, but that need not make it impossible.
This shows yet more problems with the popular usage of the word “disorder.” Perhaps it is time to abandon the word altogether.