The Neurobiology of Sleep

There is quite a bit going on, and I'm not always capable of doing it all with one hand behind my back, especially when sleep-deprived. One of my colleagues had a baby recently, and she asked us when our kids began sleeping through the night. Another colleague, mother of three, said "never." And to some extent it is true - sure, the newborn every 3-4 hour torture phase passes, but then you get teething, and then nightmares/fear of the dark, and beyond that I wouldn't know from personal experience, but let's just say I've got one teething and one afraid of the dark, and I am thoroughly sleep-deprived.

Which wouldn't bother me too much, except that I put on a couple of pounds (darn cortisol) and vanity is a failing of mine. From medical training I'm an old hand at sleep deprivation. Now there are actually laws against the amount of work I did back in the day. Not that I disagree with those laws as a rule - I happened to do my psych consultation service on some of the same floors where the medicine interns wore EEG leads post-call (meaning after being up all day and all night, and rounding on patients in the morning) as part of a study on resident work hours. Turns out they spent a good deal of those post-morning rounds in phase I sleep. Since I'm sure most of you don't care for sleepwalking doctors in charge of communicating vital information about your admission and hospital course, I imagine you are probably in favor of laws controlling medical residents' work hours. You might be surprised by the response within the medical community. They want residents to suffer - they think it makes better doctors. And working crazy hours does build confidence and experience faster than anything. However, I hope it is not terribly controversial to suggest that we not make new doctors at the expense of safety (too much).

Well, sleep! Which I am sorely missing. I'm not young anymore, after all. I have at hand a basic reference for the neurobiology of the circadian rhythms. It is from the supplements from the Journal of Clinical Psychiatry (which I normally throw away, as they are generally thinly-gilded advertisements for the pharmaceutical industry. And this supplement is no different- it came out in 2005, which was right about when Ambien CR and Rozerem were being marketed, and if you read the entire supplement you get statements about how cheap, long-used sleeping pills aren't FDA approved (of course they are not - they were generic long before the current FDA approval process, and who would bother to spend the several gazillion dollars needed for the current FDA approval process for generic medicines?) But, pharmaceutical advertising notwithstanding, the paper I'm referencing is rather too basic to be dangerous, and it is free (with free registration) for everyone, so let's call it egalitarian.


The Human Circadian System and Normal and Disordered Sleep

The neurobiology of sleep is nearly as simple as light and dark. We have, in our brains, right in the middle a bit above our eyes, a little area called the superchiasmatic nucleus. It is an area of about 10,000 neurons that runs our circadian rhythms. Without light stimulus, it tends to fire in a rhythmic pattern in a cycle of a bit more than 24 hours. Light stimulus will reign it in and keep it correlated with our natural light/dark cycles.

Light hits the retina of our eyeballs, which then sends a signal up through our optic nerve directly to the superchiasmatic nucleus. There a dimer of two proteins is made in response to the light - CLOCK+BMAL1 (let me suggest that the scientist who named the CLOCK protein was a tad more poetic than the namer of BMAL1). BMAL1/CLOCK starts the day running by binding the per/cry promoter regions in the nucleus, leading to the creation of the PER/CRY protein complex (there are in fact 2 pers and 3 crys, but let's just call it per/cry for simplicity's sake.)

From here, PERs and CRYs have to bond with other PERs and CRYs to form dimers, or else they are broken down very readily. The dimer of PER/CRY is then translocated from the nucleus to the cytoplasm for daytime cellular activity. All right! Throughout the night, when no new PERs or CRYs are made, the dimers already in the cytoplasm slowly degrade, until light comes again and more new ones are made. Thus the cycle of life and light and dark. That, in a nutshell, is the circadian rhythm.

Under conditions of sleep deprivation, our circadian clock will keep us from getting too far out of whack from light/dark cycles. We will readily sleep in the wee hours of the morning (and in fact major industrial accidents such as the Exxon Valdez crash and the Three Mile Island radiation leak happened at 3-4 am - at least according to my book on chaos and sleep), and yet even if sleep deprived and given ample opportunity, we have a hard time sleeping during the "forbidden zones" of 9am and 9pm.

And what of that famous pineal gland hormone, melatonin? It is secreted in response to dark, and light stimulus (such as checking the twitter feed on the ipad at 3am, or turning the hall light on to quell the 3 y/o's fear of the dark) will diminish melatonin secretion immediately. The human superchiasmatic nucleus (SCN) has a bunch of melatonin receptors (there are very few in the human brain outside this region). Melatonin signals the SCN to cool it and settle down for the night. Melatonin seems to sharpen the natural SCN response to light and dark. Without proper melatonin signaling, light and dark signals to sleep or wake up are attenuated, leading to night wakefulness and daytime sleepiness.

So what to do if you have insomnia? The first thing is to cut out any late night retina stimulators - like TV or internet. Music or low-light reading is probably okay. Our retinas were never designed for HDTV at midnight streaming so much signal straight into our brains (I'm reminded of when my oldest as a baby was awake at night, and my husband, trying to spare me, dutifully took her downstairs and began watching "Three Kings" with her at 2am - well, that Iraqi sun on the big screen TV at 2am did nothing for our baby's ability to sleep, and she was WIRED for several hours.) Do not turn on the lights for a midnight visit to the refrigerator or the restroom. I'm not entirely convinced that complete, black darkness is necessary given our ancestral propensity to sleep under the moon and stars, but as little light as possible is likely ideal.

Get rid of the clocks, too. I don't use an alarm and haven't for many years. If you need an alarm to wake up, you aren't getting nearly enough sleep in the first place.
There is an online resource called "CBTforInsomnia.com" that has an inexpensive program to help with insomnia (I have no connections to this resource and receive no money from promoting it).

Failing that, if you have a condition such as depression, anxiety, or bipolar disorder, sleep is exceedingly important. I'll focus more on these individual conditions in separate blog posts. In these cases I often feel it is prudent to prescribe sleep medicines (in fact, with an escalating mania, especially a psychotic mania, sleep medicines will nip it in the bud as quickly as anything else), but it is obviously not the ideal and is not a long term solution.

My own preferred sleep remedy is magnesium supplementation. I take magnesium oxide (low bioavailability, but easy to find at any drugstore) 250-500mg (depending on how many nights in a row I forget to take it), a lower dose of magnesium citrate, or a low dose of magnesium gel meant to spread on the skin, and I sleep well, right up until one of the children wakes me up. Maybe I should coat them with magnesium too.