Suppose I want to improve performance of my washing machine. Ways I might do this fall into three categories:
1. Supply missing inputs. It needs water, soap, and electricity. If any one of them is missing, I can greatly improve performance by adding it — by plugging the machine in, for example. These changes are easy because water, soap, and electricity are easy to get.
2. Replace broken parts. This will also greatly improve performance. These changes are very difficult unless I am a washing machine repairman.
3. Everything else. To improve performance any other way will be difficult and any improvements will be small. These other methods of improvement — such as putting special disks into the wash — are also likely to be dangerous.
All complex machines are like this. What I call the Washing Machine Principle says that humans are also like this. This means that non-transplant attempts to improve human well-being fall into two clusters: 1. Easy, safe, and highly effective. 2. Difficult, dangerous, and only slightly effective.
Some simple examples:
- Vitamins. If you have a deficiency disease, getting more of the right vitamin will cure you easily, safely, and rapidly. They supply a missing input.
- Antidepressants. They are dangerous, difficult to make and obtain, and don’t work very well. In controlled studies, they do only slightly better than placebos. Patients typically must try several to find one that works. They don’t supply a missing input.
- The mirror treatment for certain neurological conditions that Atul Gawande recently described:
[The patient’s] left hand felt cartoonishly large—at least twice its actual size. He developed a constant burning pain along an inch-wide ribbon extending from the left side of his neck all the way down his arm. And an itch crept up and down along the same band, which no amount of scratching would relieve. . . . [These symptoms had lasted 11 years. Gawande suggested trying the mirror treatment.] After a couple of weeks, his hand returned to feeling normal in size all day long. The mirror also provided the first effective treatment he has had for the flares of itch and pain.
The mirror treatment is cheap, safe, and, in this case, highly effective. Clearly it supplies a missing input.
To be continued.
Certainly antidepressants have the potential for harmful side-effects, but if the problem is caused by low serotonin levels, would not an SSRI supply (indirectly by inhibiting reuptake) a missing input? Since depression has so many causes and these causes are very hard to observe, it is not surprising that they would not do much better than placebos *on average*. But that doesn’t mean that they couldn’t work remarkably well for those whose depression is actually caused by low serotonin levels. I know several people who were very effectively treated by antidepressants of various types, but almost all of them required trial periods of at least two drugs, whether they be SSRIs, SNRIs, and/or NDRIs. A drug that supplies an input that’s not actually the missing one will of course not work. So of those patients who were not affected in the trials, how many do you think might have eventually found a drug that works? It’s not without risk, but as you’ve said, neither is flax oil (for pregnant women), or vitamins (for those who don’t have a deficiency).
If there were a trial that found that flax oil on average has no statistically significant effect, I would hope you would jump all over that study. Concluding that it doesn’t help *anyone simply because it doesn’t significantly effect *everyone is the kind of error that I’ve learned from you and Dierdre McCloskey to avoid (that’s why I commented a while back that you two have a lot in common despite your disagreements over gender identity). But I suspect that your skepticism of these medications disinclines you to come to this conclusion when the subject of the trial is antidepressants. What do you think?
I agree with Matthew. Saying antidepressants don’t work is sort of like saying vitamins don’t work, because zinc, copper, magnesium, iron, and potassium supplements don’t help your calcium deficiency. Antidepressants are a broad category of drugs treating a broad spectrum of symptoms. Of course, on average, they’ll be a wash.
Serotonin is not an input. By “input” I mean what we take in from the environment — what we eat, see, hear, etc. As for safety, vitamins are much much safer than antidepressants — that was my point.
I’m not saying “antidepressants don’t work” — I’m saying they work much worse than vitamins when the vitamins are used to treat a deficiency disease, e.g., Vitamin C is used to treat scurvy.
A comment that someone else was unable to post:
Matthew’s argument presumes that, for some people, “depression is actually caused by low serotonin levels.” Even strong advocates of psychopharmacology do not agree. For example, Ellen Frank notes that, “It may be that the biological changes produced by antidepressants can lead to changes in mood, concentration, and interest by setting the stage for circadian reentrainment.” (see Treating Bipolar Disorder, page 23)
Jump-starting the brain’s machinery may “workâ€; it doesn’t follow that jump-starting is the best way to operate the machine.
Antidepressants might work at treating (some of) the symptoms, but they don’t (at least in the large majority of cases) solve the cause of depression. The cause of depression is usually a basic human need being unfulfilled. Vitamins do address the cause of certain diseases – basic nutritional needs.
Low serotonin levels, properly understood, aren’t causes of depression but effects. Saying depression is caused by low serotonin levels is like saying that the washer malfunction is caused by some lever that causes the water to enter the tank not flipping on – and ignoring the fact that the lever isn’t moving because there’s no electricity being supplied to the machine.
I’m not saying “antidepressants don’t work” — I’m saying they work much worse than vitamins when the vitamins are used to treat a deficiency disease, e.g., Vitamin C is used to treat scurvy.
Comparing vitamins to treat deficiencies with anti-depressants is not entirely fair. Clearly, anti-depressants are non-optimal and treat the symptoms of depression, not the actual cause (sadly just like 99% of modern medicine), but they can be thought of as a temporary patch on a flat bicycle tire. Yes, the tire still has a hole from the thorn. Yes, the patch only covers the hole, and therefore the patch has not addressed the underlying cause i.e. the hole still exists i.e. you would prefer to ride a bike with your bicycle tube wholly intact than one with a patch…but if all you got is a patch, i.e. you can only treat the symptom, and no new bicycle tube exists, you are still a helluva lot better off with the patch than without.
—————–
BTW I am skeptical of the scurvy and vitamin C via fruit and vegetable hypothesis (yes, I know that makes me sound insane), for the following reasons:
1 – Poor absorption of antioxidants provided by fruit and vegetables.
2 – Antioxidants when tested in a lab most likely have very different effects when ingested.
3 – Antioxidants may not be beneficial, but may actually be deleterious to health. See here:
https://barrygroves.blogspot.com/2008/10/eating-fruit-may-increase-heart-attack.html
4 – Our requirements for vitamin C are probably very low.
5 – Vitamin C is required to digest carbohydrates, grain does not provide vitamin C. Less carbohydrates, less need for vitamin C. BTW according to Barry Groves of ‘Trick and Treat’, only those British sailors who were eating the ships biscuits were prone to scurvy.
The Masai, Inuit and Samburu, who eat no or very little vegetable matter do not suffer from scurvy. So, where are they getting their vitamin C from?
It appears as if organ meats such as liver are a better source of vitamin C as they provide more vitamin C and are more readily absorbed by the body.
BTW this was referenced on this very blog here:
https://sethroberts.org/2008/07/21/errors-in-the-queen-of-fats/