Smoking as a Costly Signal

  1. Smoking as a Costly Signal
    Signalling is the idea that in a population of individuals, some will send out signals as a sign of their quality (e.g. fitness as a mate) and some will receive signals to appraise potential mates. Senders benefit from having their signals accepted whereas receivers benefit from accepting signals from fit individuals. Say there is a poor signal and a good signal (i.e. the latter has a better chance of being accepted). Senders will want to send the good signal if they can. Costly signalling implies that they have to pay a greater cost for sending the better signal. Importantly, lower quality individuals pay a greater cost than higher quality individuals. Therefore there will tend to be an efficient separation such that receivers can be fairly confident that by accepting a costly signal they will be accepting a higher quality mate.In the animal kingdom, the peacock’s tail is a great example of costly signalling. The larger the tail, the more difficult it is for the peacock to fly. It seems surprising then that peacocks would evolve to have these cumbersome tails: surely they aren’t adaptive? Amotz Zahavi proposed the Handicap Principle to explain this: a male peacock with a large tail is signalling to peahens that he is so fit that he can still evade predators even with the handicap of a large tail (apologies for the terminology).Yesterday I asked myself why the dramatic increase in the price of cigarettes (at least in my part of the world) seemed to coincide with a dramatic decrease in their ‘coolness’. I thought that if I looked at the phenomenon from a costly signalling perspective then surely an increase in the cost of the signal should allow higher quality senders to more effectively separate themselves from lower quality ones.

    A quick Google search returned the paper linked above. This doesn’t match my intuition for where I’m from but Dewitte (writing from Belgium) claims that adolescent smoking has become more popular and that the reason for this is in fact, costly signalling. Adolescents who are healthier (at least before they start smoking!) can signal their fitness by adopting this habit, effectively saying “I’m so healthy that I can afford to smoke”. Note the similarity to the peacocks. Although everyone can choose to smoke, less healthy people (say someone with asthma) pays a higher cost by smoking such that receivers can reliably infer that anyone who smokes doesn’t have underlying health problems (such that they are more attractive).


On Being a Nerd

In a certain subset of my friend group, the term ‘nerd’ is unanimously considered a term of endearment. This happens to be my own default so it came as a surprise when I was recently reminded that this is not the opinion of most people.

Last week one of the people in that aforementioned subset recounted how after he had described an acquaintance as a ‘nerd’, they confronted him and said that they were quite hurt by his comment. My friend was rather taken aback (as I would be in that situation, and explained that to him, calling someone a nerd is virtually the highest compliment he can pay.

Why the difference in opinions? Is it that the two parties have different conceptions of what being a nerd actually entails or do they agree on the definition but differ in terms of the value they place on those traits?

I can’t imagine that their definitions of a nerd would be grossly different. If they were presented with ten people and had to pick out the nerd then I’d wager that they would both look for the same features and would almost certainly pick out the same person. It’s a stereotype for a reason after all.

It seems much more likely that my friend just has a rosier opinion on the actual traits of nerds. Why might this be?

Well, what makes a nerd? In general, a nerd is someone who places a higher value on being intelligent (and on being seen to be intelligent) than most people do. They also tend to prioritise their own interests (which are typically of a cerebral nature) over the attainment of popularity amongst their peers. This results in the stereotypical lack of social skills.

So if my friend calls me a nerd I take it as a compliment because I think “Hey, he must think that I’m smart and that I’m independent enough to do my own thing”. Whereas when my friend called their acquaintance a nerd, that person thought something along the lines of “Why does he think I’m not popular?”

Thankfully this opinion seems to be on the wane. There’s no doubt that society as a whole is more accepting of nerdiness and everything that goes with it nowadays than say, a decade ago. Part of this shift can be attributed to the success of the Silicon Valley enterprises run by people who would have been mocked as computer geeks when in school.

Not only are these people becoming incredibly wealthy, they are also influencing people’s conceptions of what it means to be cool. We can now draw a Venn diagram showing an intersection between nerdiness and coolness and not expect people to scoff at it.

Additionally, pop culture has more positive depictions of nerds nowadays. Granted, shows like the Big Bang Theory still portray nerds as socially inept and often seem to be laughing at them rather than with them but at least they don’t treat them as boring. That’s a big change from Ross Geller in Friends who consistently put everyone to sleep as soon as he started talking about his passion for dinosaurs.

And if you really want proof of how the nerd’s philosophy has permeated throughout pop culture, just listen to Big Sean in the song Wanna Be Cool where he basically raps a nerd credo:

Rocking pink Polo’s, shit ain’t even fit me,
Looking for the inspiration that’s already in me,
All the confidence I was trying to buy myself,
If you don’t like me, fuck it, I’ll be by myself,
Spent all this time for you to say I’m fine,
I really should have spent it trying to find myself.

Continue reading

Propagating Urges

The New York Times recently ran a piece entitled The Happiness Code which I came across because several people I follow on Facebook posted a link to it, including one of the interviewees. In the feature, the writer Jennifer Kahn describes her experience at a “self-help workshop” that promotes rationality as a way of improving one’s life.

The organisation that runs the workshops, the Centre for Applied Rationality (CFAR) is an organisation that I’ve been familiar with for a while. One of its founders, Julia Galef, runs the excellent Rationally Speaking Podcast which I recommend everyone checks out.

I’m not going to discuss the article in depth; I’d recommend you read it for yourself instead. I do want to talk about one of the techniques mentioned in it though, that of ‘propagating urges’.

Before discussing what exactly the technique involves, let’s look at what it intends to solve. I was about to write “Have you ever found yourself doing something that you know is a waste of time but lacking the motivation to do what you know you should do?” but I guess I should really be asking how many times a day you experience that. Not because I think you’re lazy but because that feeling is a universal human experience.

This is the result of a phenomenon known as hyperbolic discounting. In other words, we place greater value on rewards that we don’t have to wait a long time for than rewards that are far in the future. This is not simply a matter of us assigning a lower probability to the latter, instead the waiting around actually subtracts from how much we value the reward itself. The major downside to this is that our present-self makes decisions that our future-selves will regret.

If you’re like me then you’ll also have found yourself wishing for some easy way to jolt you out of this reverie whenever you find yourself in it, to give yourself a mental kick in the backside if you will.

This is where CFAR and their ‘propagating urges’ come in. The idea is simple: when you’re doing whatever it is that you really want to be doing, reinforce that with a fist-pump and by exclaiming “YES!” like an athlete celebrating a victory. They also recommend that you picture yourself succeeding in the goal that you’re working towards (again this sounds similar to what a lot of athletes do at the advice of their sports psychologists).

For example, Kahn describes how she gets trapped checking her e-mails until it becomes too late to go to the gym. Personally I often find it difficult to find the time and motivation to write a blog-post because unlike say, a project for my course, there are no ramifications if I fail to write one. CFAR would advise me to start making the aforementioned “victory gesture” whenever I actually do sit-down to write a post and when I finish it so that my brain gets a pleasant sensation and grows to look forward to the writing instead of shying away from it.

In essence, instead of trying to force my brain to value the far-future reward more than the near-future one (which conflicts with our tendency for hyperbolic discounting), I should train my brain to develop urges to work on the steps towards that far-future goal. Those urges will (hopefully) then propagate down the temporal chain until I achieve my objective.

To be honest, the idea of the victory gesture is a bit much for me. I’m willing to try it but I’m also going to experiment to see if there are any alternatives I can use. However, it’s important that they be equally fast because the whole idea is that the action and the reward have to be close in time to one another so that our brains learn to associate them.

Faster and Safer Treatments for Depression

I previously wrote about Selective Serotonin Reuptake Inhibitors (SSRIs), the most commonly prescribed antidepressants. In this post, I want to discuss a novel class of compounds that may be able to treat depression faster and with fewer side effects.

Why do we need faster-acting anti-depressants?
One of the main problems with SSRIs is the time it takes for them to work (3-8 weeks). As discussed by Healy and Aldred, patients are at increased risk of suicide during this period. This is because while severely depressed people have suicidal thoughts, they generally also lack the motivation to act on them. Problems arise if they start to take an antidepressant that boosts their motivation without alleviating their depressive symptoms.

Do we have any promising leads?
You might have heard of ketamine. Originally developed for use as an anaesthetic, it is now a fairly common street drug (under the name of Special K) due to its hallucinogenic properties.

Interestingly, when people with depression were admitted to hospital and put under anaesthesia using ketamine, they often reported a dramatic improvement of their depressive symptoms.

Remember that the conventional treatments for depression work slowly so this rapid improvement (within 2 hours!) was really surprising.  To quote Thomas Insel, the director of the National Institute of Mental Health: “this is the first report of any medication or other treatment that results in such a pronounced, rapid, prolonged response with a single dose”.

Now when scientists hear something like that, their ears perk up immediately and they start asking questions. Why did these patients report such a benefit? Was it the ketamine or something else? If it was the ketamine then what is it about that drug that produces the effect? Can we use ketamine to treat depression? If not, why not and can we make a similar drug that could be used?

What did they find?
Well it turned out that yes, it was the ketamine. When they tried to figure out its mechanism they discovered that the effects are mediated by the drug’s actions on NMDA receptors in the brain. This is a very different mechanism to most modern antidepressants.

The NMDA receptor is an ionotropic glutamate receptor. This means that when glutamate (the most common neurotransmitter in the human body) binds to the receptor, a channel opens up to allow ions flow across the cell membrane. Ionotropic receptors quickly convert the chemical signal of the neurotransmitter into the electrical signal of an action potential.

Ketamine is an antagonist at the NMDA receptor, i.e. it binds to the receptor in such a way that glutamate can no longer access its binding site. Without glutamate the channel remains closed and the electrical signal cannot be propagated.

Why does ketamine help in depression?
It is thought that ketamine’s effects on depression are a result of it increasing the activity of neural circuits in our forebrain. According to Thompson et al, ketamine “suppresses NMDAR-mediated excitation of inhibitory interneurons” which leads to “a mild disinhibition of the neuronal population”.

Why does ketamine act faster than SSRIs?
Our mood is regulated by a chain of biochemical reactions. Conventional antidepressants act early on in this chain so it takes a while for their effects to be passed down. In contrast, ketamine comes in near the end of the chain so its actions are more immediately apparent.

So what does this have to do with that novel class of antidepressants from earlier on?  
Well, unfortunately ketamine is not a suitable treatment for depression because of the whole host of side effects that come along with it. In light of this, scientists have attempted to develop new drugs that borrow from ketamine’s good points but leave out its disadvantages.

Recently, University of Maryland researchers announced their discovery of “compounds that could successfully treat depression in less than 24 hours while minimizing side effects”.  

They began their study by looking for drugs capable of increasing neuronal activity, like ketamine. The clever part is that instead of boosting excitatory signals (which can be dangerous) they decided to try reducing inhibitory signals, as this would end up having the same net effect.  

How did they do this?
The University of Maryland researchers developed drugs known as GABA-NAMs that reduce the effects of GABA, the main neurotransmitter involved in inhibitory signalling. The GABA-NAMs prevent this chemical messenger from binding to its receptor. Since there’s less inhibition, neuronal activity should increase.

Were these GABA-NAMs effective?
Tests in rodents were used to investigate the efficacy of the compounds. We know that healthy rats prefer to drink water containing sugar and that they enjoy social interaction. In contrast, depressed rats lose their preference for both of these activities. This is analogous to a characteristic symptom of depression in humans, known as anhedonia, an inability to feel pleasure.

The tests revealed that the drugs reversed signs of depression within 24 hours (compared to the weeks it would take for an SSRI). The GABA-NAMs also appeared to be safer because they only affect the regions of the brain that are involved in regulation of mood.

So are these drugs the next big thing?
Obviously it remains to be seen if the compounds will be as effective and as safe in humans but the study’s author, Scott Thompson is certainly optimistic: “If these compounds can quickly provide relief of the symptoms of human depression, such as suicidal thinking, it could revolutionize the way patients are treated.”

For more information, the original paper can be found here.

Selective Serotonin Reuptake Inhibitors (SSRIs)

Disclaimer: I am not a doctor and this is not medical advice. I am writing this as a knowledge synthesis for my own benefit.

Depression is the most common psychiatric problem for which people seek help and is projected to become the second leading cause of disability in the world by 2020.

Currently, the most popular treatments for depression are a class of drugs known as the Selective Serotonin Reuptake Inhibitors (SSRIs).
Examples include escitalopram which is found in Lexapro and fluoxetine which is found in Prozac (fluoxetine has a slightly different mechanism but I won’t go into that here).

Do SSRIs even work?
They certainly aren’t miracle drugs but the evidence shows that they are better than placebo.

How well they work is another matter, one which is still being debated in the literature.

We know that a benefit exists because there is a statistically significant difference between drug and placebo. What people can’t agree on is whether or not this benefit is clinically significant i.e. will SSRIs help patients enough to justify the side effects?

Irving Kirsch states that “there is a strong therapeutic response to antidepressant medication. But the response to placebo is almost as strong”. In another paper, he claims that 80% of the improvement observed on treatment is attributable to placebo. He then puts this in context: “the effect of placebo on pain is estimated to be about 50% of the response to pain medication”.

It has been shown that people who respond better to placebo will also respond better to SSRIs than those patients who get no benefit from placebo, suggesting that some people are just more likely than others to respond to treatment, regardless of what the treatment is.

Some scientists, including Kirsch, argue that antidepressants do more harm than good so they should only be used in the most severe cases. The problem with this is that even if true, many patients with depression would happily accept that harm in exchange for relief of their symptoms.  

How do SSRIs work?
I’m going to deal with this in two parts. Part 1 will explain the mechanism by which the drugs act. Part 2 looks more at what we know about how they affect depression.

As the name implies, SSRIs inhibit the reuptake of serotonin. Serotonin is a neurotransmitter, meaning that it transmits messages between neurons across a junction known as the synapse. Neurotransmission includes the following steps:

  • Release: A signal causes the presynaptic neuron to release the neurotransmitter which diffuses across the synaptic cleft (the gap between two neurons).
  • Receptor Activation: Receptors on the postsynaptic neuron recognise the neurotransmitter such that the signal is propagated.
  • Reuptake: Afterwards, the neurotransmitters can either be enzymatically degraded or they can be taken back up into the presynaptic neuron.

There are over one hundred different known neurotransmitters and they are all involved in various different processes. Serotonin is known to regulate emotion, reproduction, attention, sleep, learning and memory.

How SSRIs affect depression 
So why do we want to inhibit the reuptake of serotonin?
The conventional explanation is that depression is caused by a deficiency in serotonin so the more serotonin that there is acting on the postsynaptic neuron, the better. This can be achieved by inhibiting its reuptake. SSRIs ensure that serotonin hangs around for longer in the synaptic cleft such that its action on the receptors is maintained. An analogy would be a boss convincing her employees to work overtime in order to compensate for falling profits.

While this explanation still gets trotted out, it really doesn’t hold water. The fact that SSRIs work is the main evidence in favour of it but in actuality, some drugs (e.g. tianeptine) which increase serotonin reuptake (the opposite effect to SSRIs) nevertheless help with treating depression.

The honest answer then is that we don’t know. One of the more in vogue explanations at the moment is that depression is related to inflammation but this isn’t much more than an admission of our ignorance since inflammation is such a catch-all term.

It has also been suggested that SSRIs strengthen and heal the brain so that it becomes less susceptible to depression but if anything, the opposite appears to be true (See Increased Risk of Relapse later). The reason that this hypothesis was first put forward was because some studies indicated that antidepressants stimulated neurogenesis, the growth of new neurons.

However, the method that they used to test this can’t distinguish between neurogenesis and neuronal death. It turns out that antidepressants actually cause neuronal damage so if neurogenesis does occur, it is probably the brain trying to make up for the harm done by SSRIs.

What are the problems associated with SSRIs?
While these anti-depressants are widely prescribed, they have a number of limitations. These include:

  • Delayed action. It can often take several weeks for the drugs to start working. To compound matters, symptoms (including suicidal ideation) often get worse initially.
  • Side effects. These are generally milder than for other antidepressants but since serotonin is involved in the regulation of so many physiological processes, the range of side effects is quite broad. Sexual dysfunction and weight gain are particularly common while both drowsiness and insomnia have been known to occur in different patients.
  • Increased risk of relapse. Antidepressants get less effective with prolonged use so ideally we’d like to discontinue therapy once the patient has stabilised. If the risk of relapse is substantial however, discontinuation could do more harm than good.
    Getting concrete figures on the size of the risk is tricky but Andrews et al found that the 3 month risk of relapse increased from 4% for patients who had been given a placebo to 43.3% for SSRIs and to 61.8% for fluoxetine. Unfortunately they don’t discuss how clinically meaningful this is.
  • Difficulty concentrating. You know how a computer with low RAM can’t do many things at once? Depression causes a similar problem because serotonin is involved with working memory. In theory, we would expect SSRIs to alleviate this but as we’ve seen so many times, the theory doesn’t translate. Patients might get a short respite but their attention spans seem to get worse in the long-term.

Are SSRIs only effective in severe cases?
You may have heard that SSRIs are ineffective in patients with mild depression. For example, Kirsch found that “efficacy reaches clinical significance only in trials involving the most extremely depressed patients, and that this pattern is due to a decrease in the response to placebo rather than an increase in the response to medication”.

However, this may have more to do with the quantification of efficacy than the drugs themselves. Essentially, it’s easier to quantify an improvement in patients with major depression than those with a less severe form.

When clinicians want to grade a patient’s depression, they often use the Hamilton Depression (HAM-D) scale. The minimum score on the HAM-D scale is 0 (no depression) while the maximum score is 52. It seems to make sense then that researchers would consider the size of the reduction in a patient’s HAM-D score on treatment to be indicative of how effective the medication was. Unfortunately, this isn’t necessarily the case.

Consider a drug which relieves depression by 20%. If a patient with a HAM-D score of 25 (very severe depression) is given this drug, then their score should decline to 20 (severe depression). That’s pretty impressive. However, if the same drug is given to someone with a HAM-D score of 12 then their score only falls to 9.6. Both 12 and 9.6 are in the same category, mild depression.
It should be clear then that the results will always look more impressive when the drug is given to a more severely depressed patient.

On the other hand, larger fluctuations in the HAM-D score can occur quite naturally. Kirsch writes that “a six-point difference can be obtained just by changes in sleep patterns, with no change in any other symptom of depression”. He also points out that the drug-placebo difference was below the level of clinical significance even when he just analysed patients with very severe depression.

The problem with this is that ‘clinical significance’ was defined completely arbitrarily by guidelines such as NICE in the UK who later admitted that their choice was based more on picking a round number than picking one that was meaningful.  

I can’t finish this without highlighting one critical point: SSRIs do work but that doesn’t mean that they are the best option for every patient. Goodwin et al, writing in relation to the use of the drugs in bipolar disorder, emphasise the gap that exists between physicians who are happy to prescribe anti-depressants and researchers who “agonize over whether antidepressants should be offered at all, at least as monotherapy”. 

What this all suggests is that SSRIs probably shouldn’t be first-line treatment for mild depression but they can be a valuable therapy for patients with moderate depression. Physicians should remember that prescribing an SSRI straight away might not be the safest option. Patients should remember that multiple options exist; there are other drugs and other therapies. Cognitive Behavioural Therapy (CBT) in particular is a well-tolerated treatment.

How Should Rationalists Approach Death

While I certainly agree with a lot of what Yudkowsky has to say (being angry about death and not needing to find a comforting explanation for example), I do wonder if cryogenics is really an optimal solution. This may seem like a whimsical opposition but I would honestly be worried about who would maintain control of my cryogenically frozen body given that it is impossible to know how long it will take before I could feasibly be thawed out. Say I was Fry from Futurama and I was preserved for 1000 years, who’s to say my body isn’t going to be in the hands of some nefarious person or organisation? I feel like there’s definitely a novel that could be written about the possibility of a dystopian future where some crazed lunatic has gained control of the bodies of everyone who elected to be frozen and then reanimates them into a manufactured version of Hell.

The music of Pi

The music of Pi

We all know that π is an irrational number, meaning that it is a real number which cannot be written as a fraction.

This website, π , takes the first 100,000 digits of π and assigns each number between 2 and 9 a musical note. When a 1 or a 0 pops up, this changes the display to day and night respectively.

The actual sound is almost reminiscent of a Bjork song and is definitely worth checking out.