I have no idea.
Seems like a question for scienceofeds!
What an astute observation anon! I will respond to this in more detail sometime today.
Okay, I cheated. I outsourced this question to a friend of mine because I’m busy and because she’s more qualified than I am. I’ve been trying to a while to get her to write blog posts for SEDs with no luck.
Here’s her answer (with my interjections in square brackets; all the bolding/italicizing/underlining and linking is also my doing)
5-HT = serotonin and 5-HT(1/2)R = serotonin receptor (that is, what serotonin binds to in the brain).
I’m going to put her concluding paragraph in the beginning:
The simple answer in regards to depression and AN is that both illnesses are likely not just too little or too much serotonin and we really only have pieces of the puzzle in terms of understanding both illnesses. It’s too soon to say whether we can apply a completely different neurochemical model to AN vs. depression - for this we need better powered studies with many more people (both recovered and ill) and I think we also need to stop pooling AN-R and AN-BP, as we don’t know if these have completely similar neurochemical bases - However, this would be difficult given the cross-over rate between subtypes.
[Not only is focusing just on serotonin very simplistic, as there’s more to these illnesses than just serotonin and there are many serotonin receptors in different brain regions that have opposite effects on how neuronal activity — really, receptors, not the actual chemicals, are the important pieces of the puzzles — but the same behaviours or states (e.g., depression) could have different causes/causal factors. This would make sense, given that SSRIs don’t work for everyone with depression (LOL BY FAR). There’s no reason to think that just because person A and person B are both diagnosed with AN or BN or depression, that the underlying neurochemistry is the same. Very important point: There are also NO studies in AN and BN that look at the neurochemistry BEFORE the development of the ED, so the idea that these changes / differences are causal is a HUGE assumption.]
There is some evidence that AN is related to an excess of serotonin, but this is indirect (as we can’t actually measure serotonin in the brain, so anything we can do is indirect). Even with PET imaging [which is what’s used to study the levels of neurotransmitters in the brain, and the studies that have led to the hypotheses that anon mentions], at this point we have no good radioligands that are competitively displaced by serotonin which would give us an index of how much serotonin someone has in a certain state or after a certain challenge (i.e. administration of a hormone, starvation vs. recovery, etc.). There are groups working on this though and we do have a tracer like this for dopamine in which we can directly measure the amount of tracer displaced from a receptor by dopamine after a pharmacological/behavioural challenge or in a particular state of illness vs. recovery…
At the moment, one thing we can do is look at serotonin turnover, as measured by levels of its metabolite in CSF, which we assume is in equilibrium with serotonin synthesis so is an accurate measure of serotonin levels. However, this does not say anything about whether serotonin is elevated in some brain regions and not others and this is making a rather giant assumption.
In this respect, levels of 5-HIAA (a serotonin metabolite) have been found to be reduced in those with AN compared to healthy controls and elevated after recovery - supporting the hypothesis of a hypersertonergic state predisposing AN, leading to anxiety, harm avoidance, and a spiral of restriction etc. However, I’m not sure how well powered these studies were (i.e. in terms of sample size) and if they compared AN-R and AN-BP, recovered and healthy.
The second thing we can do is measure levels of various serotonin receptors (like the 5-HT1A receptor) using PET because these usually become desensitized upon repeated interaction with an agonist [something that activates the receptor] (i.e. serotonin) - so high states of serotonin should produce less 5-HT1A receptor binding as measured by PET and low states of serotonin should cause an upregulation of the 5-HT1AR, so greater 5-HT1AR levels.
There have been studies which have shown increased 5-HT1AR binding in AN as compared to health (i.e. less serotonin, anxiolytic, up-regulation of the 5-HT1AR), but some of these studies were grossly underpowered with sample sizes of 8 vs. 28 for ill and healthy participants, respectively (Galusca et al., 2008). [Underpowered = the sample size is likely too small to detect any differences that have practical implications, for one, although it is more complex than that.]
In addition, this study included 9 recovered participants who also showed increased 5-HT1AR binding, which would not be expected if people recovered from AN have more available serotonin upon recovery (i.e. we would expect receptor down-regulation). [So, not all studies agree! There’s some conflicting findings.]
Also, the information can vary greatly depending on the tracer and how the data is analysed (researchers tend to use the analysis method which gives them the best results).
Another study (Bailer et al., 2007) using a different tracer and method of analysis compared a larger sample of AN subjects - 15 (but 8 AN-R and 7 AN-BP) vs. 29 healthy controls and also found increased 5-HT1AR binding in AN vs. healthy subjects - so this does support the hypothesis of less serotonin during starvation leading to up-regulation of the 5-HT1A receptor and anxiolysis, but the jury is still out on what happens during recovery.
[Here are links to blog posts where I expand on this and talk about serotonin levels in AN, or at least, what we know/don’t know: here and here
In terms of depression and treatment of AN with SSRIs, my guess would be that since the literature points to the fact that people with acute AN have less serotonin (and also display many depressive symptoms during the acute phase of the illness), it’s not necessarily a bad thing to prescribe SSRIs to alleviate some of the depressive symptoms. However, I’m not aware of any research supporting their use once an individual is in recovery.
In terms of BN, there is evidence of elevation of the serotonin transporter (again, using PET, but with a crappy radiotracer) and the serotonin transporter reduces the amount of available serotonin. SSRIs are antagonists of the serotonin transporter so increase extracellular serotonin - therefore, it makes sense that SSRIs would be an effective treatment for BN, and at the moment, I believe SSRIs are the only evidence-based pharmacological treatment for eating disorders (BN specifically). [This is true, 60mg has been shown to be somewhat effective at reducing binge eating and especially purging in patients with BN.] Low serotonin can also affect appetite (which is why weight loss can be a side effect of some SSRIs which elevate serotonin), so this might also play into the b/p behaviours seen in BN - but again, this is an overly simplistic answer to a complex question.