Last month, Scott Alexander tried to show the evidence, pro and con, on whether we have detected a causal relationship between physical ailments and depression:

Start with From inflammation to sickness and depression [PDF], Dantzer et al (2008), who note that being sick makes you feel lousy [citation needed]. Drawing upon evolutionary psychology, they theorize this is an adaptive response to make sick people stay in bed (or cave, or wherever) so the body can focus all of its energy on healing. A lot of sickness behavior – being tired, not wanting to do anything, not eating, not wanting to hang around other people – seems kind of like mini-depression.

All of this stuff is regulated by chemicals called cytokines, which are released by immune cells that have noticed an injury or infection or something. They are often compared to a body-wide “red alert” sending the message “sickness detected, everyone to battle stations”. This response is closely linked to the idea of “inflammation”, the classic example of which is the locally infected area that has turned red and puffy. Most inflammatory cytokines handle the immune response directly, but a few of them – especially interleukin-1B and tumor necrosis factor alpha – cause this depression-like sickness behavior.

[…]

Here are some other suspicious facts about depression and inflammation:

– Exercise, good diet and sleep reduce inflammation; they also help depression.

– Stress increases inflammation and is a known trigger for depression.

– Rates of depression are increasing over time, with the condition seemingly very rare in pre-modern non-Westernized societies. This is commonly attributed to the atomization and hectic pace of modern life. But levels of inflammation are also increasing over time, probably because we have a terrible diet that disrupts the gut microbiota that are supposed to be symbioting with the immune system. Could this be another one of the things we think are social that turn out to be biological?

– SSRI antidepressants, like most medications, have about five zillion effects. One of the effects is to reduce the level of inflammatory cytokines in the body. Is it possible that this is why they work, and all of this stuff about serotonin receptors in the brain is a gigantic red herring?

– It’s always been a very curious piece of trivia that treating depression comorbid with heart disease significantly decreases your chances of dying from the heart disease. People just sort of nod their heads and say “You know, mind-body connection”. But inflammation is known to be implicated in cardiovascular disease. If treating depression is a form of lowering inflammation, this would make perfect sense.

– Rates of depression are much higher in sick people. Cancer patients are especially famous for this. No one gets too surprised here, because having cancer is hella depressing. But it’s always been interesting (to me at least) that as far as we can tell, antidepressants treat cancer-induced depression just as well as any other type. Are antidepressants just that good? Or is the link between cancer being sad and cancer causing depression only part of the story, with the other part being that the body’s immune response to cancer causes inflammatory cytokine release, which antidepressants can help manage?

– Along with cancer, depression is common in many other less immediately emotion-provoking illnesses like rheumatoid arthritis and diabetes. The common thread among these illnesses is inflammation.

– Inflammation changes the activity level of the enzyme indoleamine 2,3 dioxygenase. This enzyme produces kynurenines which interact with the NMDA receptor, a neurotransmitter receptor implicated in depression and various other psychiatric diseases (in case your first question upon learning about this pathway is the same as mine: yes, kynurenines got their name because they were first found in dog urine).

– Sometimes doctors treat diseases like hepatitis by injecting artificial cytokines to make the immune system realize the threat and ramp up into action. Cytokine administration treatments very commonly cause depression as a side effect. This depression can be treated with standard antidepressants.

– Also, it turns out we can just check and people with depression have more cytokines.

There’s also some evidence against the theory. People with depression have more cytokines, but it’s one of those wishy-washy “Well, if you get a large enough sample size, you’ll see a trend” style relationships, rather than “this one weird trick lets you infallibly produce depression”.

[…]

So in conclusion, I think the inflammatory hypothesis of depression is very likely part of the picture. Whether it’s the main part of the picture or just somewhere in the background remains to be seen, but for now it looks encouraging. Anti-inflammatory drugs do seem to treat depression, which is a point in the theory’s favor, but right now the only one that has strong evidence behind it has side effects that make it undesirable for most people. There’s a lot of room to hope that in the future researchers will learn more about exactly how this cytokine thing works and be able to design antidepressant drugs that target the appropriate cytokines directly. Until then, your best bets are the anti-inflammatory mainstays: good diet, good sleep, plenty of exercise, low stress levels, and all the other things we already know work.