The Inflamed Mind Page 13
I say “as if” because I want to stay on guard against the charge of anthropomorphism. We obviously can’t know for certain that the mouse feels depressed, or that it can conceive of its own life as more or less pleasurable than it used to be, or that it can imagine that the lives of other mice are more or less pleasurable than its own. All we know is that, under normal circumstances, given a choice between plain water and water that has been sweetened with sugar, mice, like many children, will prefer to drink the sugary liquid. We assume their behavioural preference is motivated by a rewarding sense of pleasure, as we know it is in children. We also know that when the mice are tested again after BCG vaccination, they will no longer prefer to drink sugary water. They have become behaviourally indifferent to the choice before them, and we assume that is because they have lost the hedonistic, pleasure-seeking drive to consume sugar. From their changed behaviour we can infer a change in their mental experience of pleasure, akin to the loss of pleasure or anhedonia that is a core symptom of MDD. Personally, I think this is a reasonably solid line of reasoning that minimises the translational gap between animal and human, and so is not ridiculously anthropomorphic. I can fondly imagine that Descartes himself might have agreed with me, but I can’t be sure.
It is philosophically less complicated to focus on animal experiments that tell us something about how inflammation affects the brain, indisputably part of the body machine, than on the behaviour or the inferred mind of the animal. We know that if a dose of bacterial toxin, like LPS, is injected into the bloodstream of a rat, the LPS molecules themselves will not immediately get into the brain. The BBB will keep them out. But the rat’s inflammatory response to LPS will penetrate the BBB. Cytokines released from activated macrophages in the rat’s body can transmit an inflammatory signal across the BBB that will then activate the macrophages resident in the rat’s brain.
For historical reasons, the brain’s macrophages are called microglial cells, but despite their different name, microglial cells are very similar to macrophages anywhere else in the body. They spend most of their life waiting quietly for trouble, for a local invasion of hostile agents, or for a broadcast call to arms issued by immune cells under attack in other parts of the body. When microglial cells pick up the inflammatory signals generated by the body in response to an LPS injection, they get angrier, more mobile, and they start pumping out cytokines themselves, effectively echoing or amplifying the body’s state of inflammation in the brain. And, as elsewhere in the body, the mobilisation of the brain’s microglial robocops causes collateral damage to innocent bystanders, to the nerve cells in the surrounding tissue.58
When the macrophage army is mobilised under shoot-to-kill orders, whether it is in the lung or the joints or the brain, it will always lay waste to the neighbourhood. The brain is at least spared the scarring that can follow chronic inflammation in other parts of the body - the brain doesn’t form mechanically distorting scar tissue in the same way that Mrs P’s hands were deformed by fibrous contractures around the joints of her fingers. But the brain suffers from the collateral damage of microglial activation in other ways: nerve cells are more likely to die or to shrink in size, synaptic connections between cells are more likely to be rigid than plastic, and the synaptic supply of neurotransmitters like serotonin is likely to be disrupted.
Not only can angry microglial cells can kill nerve cells in their immediate neighbourhood, they can also block the regenerative process that would form new nerve cells in their place. Less extremely, but still seriously, microglial activation can make nerve cells less adaptive, or less plastic. Nerve cells, especially the synaptic connections between nerve cells, are normally plastic. That doesn’t mean, for avoidance of doubt, that nerve cells are made of polystyrene or PVC; it means they are malleable or can be remodelled, as if they were made of plasticine. Synaptic connections can be strengthened or weakened, over the course of time, so that the most useful or frequently used connections between nerve cells become stronger and the less useful or usual connections become weaker. Freud was one of the first people to imagine this idea, although he couldn’t actually see a synapse or be certain that they existed at the time (Figs. 5 and 8). Now synaptic plasticity is known to be fundamentally important for adaptive behaviour, learning and memory. So the loss of synapses and synaptic plasticity caused by microglial activation provides a plausible explanatory link between inflammation and the memory loss, cognitive impairment and quasi-depressive behaviours observed in inflamed animals.59
Microglial activation also has disadvantageous effects on how nerve cells handle the transmitters that pass signals from one cell to another across the synaptic gap. This is especially clear for serotonin, the neurotransmitter that is targeted by SSRIs. Normally nerve cells make serotonin from a raw material called tryptophan. But the cytokines released by angry microglial cells can instruct nerve cells to use this same material to make other end-products, such as kynurenine.60 This is bad news in two ways. First, it means that there is less serotonin available for release into synapses, so the normal rhythms of serotonin signalling that are thought to be important in controlling sleep, appetite and mood will be disrupted. Second, kynurenine and many of the other molecules produced instead of serotonin are toxic. They poison nerve cells, making them over-excited and metabolically exhausted, and ultimately killing them.
The net effect of microglial activation is that serotonin signalling is disabled and usurped. Bearing in mind the theoretical importance of serotonin for depression, and for how many anti-depressant drugs are supposed to work, these effects of inflammation in animal brains could explain how, at the most fine-grained level of molecules, inflammation can cause depression. When inflammation reduces the amount of serotonin released into synapses, it is effectively pulling in the opposite direction to SSRIs, which are supposed to increase synaptic serotonin levels. This might be one reason why many patients with so-called treatment-resistant depression, who don’t respond well to treatment with SSRIs or other anti-depressants, are particularly likely to be inflamed.61
• • •
MDD in psychiatric patients, and milder depressive symptoms in the general population, are both robustly associated with increased levels of inflammatory proteins in the blood. That much seems beyond doubt from the growing volume of case-control and epidemiological studies that have been published in the last 20 years. Although a psychiatric diagnosis of MDD is conventionally incompatible with bodily disease, the association between MDD and inflammation is mechanistically compatible with the enormous number of medical patients, hiding in plain sight like Mrs P, whose depression arises in the context of an inflammatory disorder of the body.
We also now have strong evidence that inflammation can precede or anticipate depression, which is a necessary condition for inflammation to be causal. And we have increasingly good answers to the how question. We can see how a cytokine signal could get across the barrier between the body and the brain, traditionally regarded as impenetrable. In humans, we can see how even a minor inflammatory shock, like vaccination, can increase activation in regional hotspots of an emotional brain network. In animals, we can see in more detail how inflammation of the body can spread to the brain, activating the brain’s own macrophages or microglial cells, and causing collateral damage to nerve cells, synapses, and serotonin metabolism. In humans and mice, at the coarse scale of fMRI and the fine scale of cells and molecules, we can see how inflammation can cause changes in the brain that in turn cause depressive changes in our states of mind or in an animal’s behaviour.
This is the tip of an iceberg in the scientific literature and there is much more detail available for those who are interested in knowing more.4, 10, 12, 62 Yet the hardest Cartesians, the most die-hard dualists, will still not be convinced. They will complain that the evidence is not yet so extraordinarily good as to justify the extraordinary claim that the mind and the body are linked by the immune system. Indeed, to be fair, the mechanistic narrative is not yet crystal clea
r. There are many loose ends and lacunae, many gaps between what we know about animals and what we know about humans, many results that are based on a few small studies or experimental methods that have quickly become old-fashioned as the science of neuro-immunology has moved forward so fast. But this is the normal state of any rapidly progressing science. And such dramatic recent progress has already brought us to the point where the how question, if not yet entirely solved, looks like an increasingly soluble and sensible question to be asking.
Chapter 6
WHY?
We could know everything there is to know about how inflammation causes depression, and I confidently expect that we will know much more in the next few years; but we would still be left with a sense of incompleteness. There would still be something missing if all we knew was how.
We would still need to know why. Why are some depressed patients inflamed in the first place? And, more generally, why is it that the inflammatory response of the immune system, which is supposed to help us survive in a hostile world, seems to be acting against us, by making us depressed when we are inflamed?
What could make you inflamed (and depressed)?
There are several possible sources of inflammation in the body that could be relevant to depression.
One apparently obvious candidate is inflammatory disease. We now know that depression is common in patients, like Mrs P, who have a major inflammatory or auto-immune disorder such as rheumatoid arthritis, diabetes or atherosclerosis, to name but a few. However, physical disease is an unlikely explanation of the increased levels of cytokines or C-reactive protein (CRP) that have been reported in research studies of major depressive disorder (MDD). This is because, according to the official diagnostic criteria of the American Psychiatric Association, depressed patients can only have a diagnosis of MDD if they do not also have a bodily disease. Somewhat bizarrely, in my view, this means that someone like Mrs P, who ticked almost all the diagnostic boxes for depression, cannot, strictly speaking, be said to have an MDD. In clinical practice, the many people in Mrs P’s position are most likely to have their mental health symptoms either ignored or diagnosed as a case of so-called “co-morbid” depression. This label of co-morbidity means that their doctors recognise that their depression is associated with an inflammatory disorder like arthritis; and it is therefore not an MDD as psychiatrists have defined it; but neither is it caused by the same pathological process in the immune system that causes the arthritis.
Molière might have recognised co-morbidity as an example of the medical tendency to dress up the patient’s symptoms in fancy language without actually explaining where they come from. To this day, a good Cartesian doctor can use the phrase co-morbid depression as a coded way of saying to his patients “Well, you would be, wouldn’t you?” Au contraire, as we have seen, there is now evidence to suggest that a large proportion of what we’re currently calling co-morbid depression, supposedly caused by a purely mental meditation on the sad fact of being diseased, is in fact inflamed depression, mechanistically caused by the very high levels of cytokines and macrophage activity generated by major inflammatory disease.
In any case, by definition, bodily disease cannot explain where inflammation comes from in patients with MDD. What are some other plausible culprits? Or, to put it more precisely, are there any known risk factors for a psychiatric diagnosis of depression that can also drive increased inflammation?
Body fat, or adipose tissue, is inflammatory. About 60% of the cells in adipose tissue are macrophages, the robocops of the immune system, and one of the principal sources of inflammatory cytokines. Overweight or obese people, with a higher body mass index, will generally have higher blood levels of cytokines and CRP than slimmer people.62 We also know that overweight people are more likely to be depressed.63 But is this because obesity causes depression or because depression causes obesity? The causal arrow could point either way, or both ways. Depression could cause behavioural changes, like comfort eating of high-calorie foods, which lead to obesity. Or it could be the other way round. Obesity could cause depression psychologically by exposing people to stressful criticism, and self-criticism, about their physical appearance in our body-shaming culture. Or obesity could cause depression immunologically by increasing the total number of macrophages in the body and increasing the levels of cytokines in the blood. However, it is at least clear that obesity both causes inflammation and increases the risk of depression.
Age, like obesity, is both a cause of increased inflammation and a risk factor for depression. As we grow older, our bodies tend to get more inflamed. Cytokine and CRP levels increase, all other things being equal, over time. Our innate immune system is kept at a steadily increasing level of threat awareness as we grow older.64 And, as we grow older, we also tend to get more anxious and depressed. Compared to obesity, the causal relationships are a bit clearer. I think we’d all agree that ageing, or at least the passage of time, is not caused by either depression or inflammation. The clock ticks at the same rate whether you’re melancholic or not, inflamed or not. So it seems safe to say that increasing age causes increased inflammation, and increased risk of depression, not vice versa. But does this increased inflammation account for all the age-related increase in risk of depression; or is it depressing enough to be aware of your ongoing march to the grave, regardless of your blood cytokine levels? It’s not yet safe to say.
Besides age and obesity there are several other possible factors that can both increase inflammation and increase risk of depression. The body’s inflammatory status, for example, shows a marked seasonal variation, with higher blood levels of cytokines in European people enduring their winter months of November, December and January, but lower levels during the same months in Australian people enjoying a southern-hemisphere summer.65 Evidently the immune system is more prone to become inflamed in winter, presumably because the risk of flu or other infectious diseases is increased in winter, which is also a time of increased risk of depressive symptoms, especially for people with seasonal affective disorder. Is that a coincidence, or could seasonal or circadian rhythms in the immune system be driving annual or daily mood swings? We don’t know, yet.
As you can see, neuro-immunology is still too young a science to have answered all the questions it has begun to ask. But, intriguingly, one of the clearest leads to have emerged so far as a plausible source of bodily inflammation in depressed people is not a physical factor, like age or obesity or hours of daylight. It is a social factor.
Flaming stressed
Stress is both one of the most well-known, and one of the least well-understood, causes of depression. It is a familiar fact of life, which all of us probably have experienced at first or second hand, that stressful events can make people depressed. The epidemiological research confirms that this is a massive effect, especially for some kinds of stress, so-called major life events, like death of a spouse or parent or child, or loss of a job, or some other bereavement or humiliation. Your chances of becoming depressed under those circumstances are up to nine-fold greater than the background risk of depression.66 To turn it around the other way, about 80% of all episodes of depression have been preceded by a stressful life event.67 The most depressing stresses are events that involve both loss of an important relationship and social rejection. So a man who has initiated divorce proceedings against his wife will be at 10 times greater risk of depression, because of the loss of the marital relationship; but a man who is being divorced by his wife will be at 20 times greater risk of depression, because the loss of his marriage is compounded by the humiliation of being dumped.68
The impact of stress on risk for depression couldn’t be clearer. What has not been so clear is how social stress can have such a catastrophic effect on depression. There is, as always, the good Cartesian view that you would be, wouldn’t you? If your wife had gone off with somebody else, if you’d just been fired, I bet you wouldn’t be too happy either. But, as always, that isn’t scientifically illuminating, or ther
apeutically helpful. It can imply that becoming depressed in response to stress is a personal choice, or evidence of an insufficiently stoic character, or in some other way one’s own fault. The pain of the stress is compounded by the shame of a moral failure to get over it. The last 20 years have witnessed a remarkable growth in support for an alternative explanation, based on the body’s inflammatory response, rather than the mind’s introspective reflections.
One of the first clues that major life events can have an impact on the immune system was the actuarial fact that life expectancy is reduced by bereavement.69 If your wife divorces you, or you suffer some other horrible event in your life, not only will you be at very high risk of depression, you will also be more likely to develop cancer or heart disease, and your predicted lifespan will be shorter than it was before the event. We talk about dying of a broken heart as if it was merely a figure of speech but we know it is happening all around us: people are dying younger than we’d expect following the loss of a loved one. I have heard many stories of a long-married couple dying within a few weeks of each other. Haven’t we all? And a recent study confirmed that the risk of death due to a heart attack or stroke is doubled in someone who has recently been bereaved.70 The emotional and social shock of losing your life-long partner has a massively negative effect on your fitness to survive. Insurance companies know this well. It is why they offer bereavement counselling to their customers. Grief can kill you. It’s another stubborn fact that could have an immunological explanation.
We now know that a stressful life event throws a rock in the pond of the immune system, causing major changes in how all the different types of immune cell are working and interacting with each other.71, 72 The macrophages of the innate immune system, patrolling the front line of the self, are made angry or more activated by bereavement, and pump more inflammatory cytokines into the circulation.73 The excessive activity of the macrophages could inflame atherosclerotic arteries, increasing the risk of blood clots forming in the blood vessels of the heart or the brain, and thus making a heart attack or stroke more likely. The impact of social stress on the immune system is one explanation for how you could die of a broken heart.