I discuss melatonin's effects on sleep & its safety; I segue into the general benefits of sleep and the severely disrupted sleep of the modern Western world, the cost of melatonin use and the benefit (eg. enforcing regular bedtimes), followed by a basic cost-benefit analysis of melatonin concluding that the net profit is large enough to be worth giving it a try barring unusual conditions or very pessimistic safety estimates.
Full essay: http://www.gwern.net/Melatonin
Your argument essentially amounts to the following:
If all of these are true, then who wouldn't want to take it? However, you spend a lot of time on discussing point 3, but little on points 1 and 2, which are arguably the most important. How do you know that Melatonin really improves sleep quality so much? Is it just based on your personal experience (and perhaps that of other people you know)? If so, that is not convincing, as large scale randomized controlled studies are generally the only way to reliably tell if a medicine works. There are too many complicating factors like individual differences between people, the placebo effect, random fluctuation, reversion to the mean, difficulty in remembering how we felt in the past, etc. to rely on anecdotes.
Another point that your article does not address is the fact that there is a difference between a medicine having no known side effects, and a medicine ACTUALLY having no side effects. Any time that you take medicine you are taking a risk of a reaction that is unknown, or which failed to be uncovered in any studies that were done on it. For example, it is probably unknown whether a decade of Melatonin use (rather than just one or two years) causes problems of any kind. This sort of danger is unfortunately difficult to quantify, but I believe deserves at least some mention.
spend a lot of time on discussing point 3 This is by far the biggest failing of the post, it grates hugely. It's cheap, we get it.
It sounds interesting but I'm a little wary of your one line dismissal of any potential side effects without reference. To the best of my knowledge the function of sleep is still not completely understood and the long term effects of reduced sleep are not known. A suggestion to take any kind of supplement every day for the rest of your life places a fairly high bar on safety. Taking melatonin to overcome jet-lag seems very likely to be safe but I'm more wary of using it on an ongoing daily basis.
Do you have any references to support the claim that there are no long term side effects of daily use?
I did link to Wikipedia for a reason; see http://en.wikipedia.org/wiki/Melatonin#Availability_and_safety . But besides AngryParsley's link, there's
Given the unanimous results of safety in the short-term, positive results in long-term child use, the exploitation of a regular physiological process, the long track-record of melatonin use, and the lack of evidence for any long-term harm, I think I'd say the onus is on any doomsayers.
(No doubt there's a witty Eliezerism or post on the topics of negative results and burdens of proof, but offhand I can't think of it.)
It took large scale randomized studies to establish the negative health effects of vitamin supplements/antioxidants and HRT both of which appeared safe in the short term, exploited a regular physiological process and had a long track record of use. I'd want to see a large randomized study of the long term effects of melatonin use in adults to establish the long term effects of melatonin use in adults.
The Wikipedia link you give merely concludes that "evidence suggests that melatonin is safe with short-term use, three months or less". From your other links:
...The findings of this review suggest that exogenous melatonin is a relatively safe substance when used in the short term, over a period of days or weeks, and is safe at relatively high doses and in various formulations. However, the safety of exogenous melatonin when used in the long-term, over months and years, remains unclear.
There are no published long-term safety data on the use of melatonin for whatever purpose, assuming long term to mean more than 6 months of daily medication. In the light of its physiological role in animals, the potential deleterious effects include inhibition of reproductive function, delayed
These supplements are unlikely to help a balanced diet, there is little evidence they do, and there are studies which have indicated actual harm from the consumption of multivitamins
What I'd really like to see is a study comparing unhealthy diets plus multivitamins to just the unhealthy diet.
Without having looked at the actual studies in detail, it seems that even several of the "no benefit" studies report multivitamins to be beneficial when one does have an otherwise unhealthy diet.
...Similarly, a 2006 report for the United States Department of Health and Human Services concluded that "regular supplementation with a single nutrient or a mixture of nutrients for years has no significant benefits in the primary prevention of cancer, cardiovascular disease, cataract, age-related macular degeneration or cognitive decline."[16] However, the report noted that multivitamins have beneficial effects in people with poor nutritional status, vitamin D and calcium can help prevent fractures in older people, and that zinc and antioxidants can help prevent age-related macular degeneration in people at a high risk of developing this disease.
In 2007 the United Kingdom Food Standards Agency published an updated set of recommendations for eating a healthy diet.[17] The recommendations stated that pregnant women should take extra folic acid and iron and that older people might need extra vitamin D and iron. However, the report advised that "Vitamin and m
Pfff, 5959% is nothing. You can get a banana for 15 cents and it'll keep you from dying of starvation. So let's say you value the next 50 years of your life at $7.25 an hour... That's a return of 28419037481114.5%!
Seriously do you have a good link for where to buy melatonin?
I tried it for a few weeks and didn't notice any major difference. I think I'll try again on this recommendation. Perhaps my endogenous melatonin is already sufficient, or I was a lazy self-monitor.
I would summarize:
(1) In your personal experience, 1.5mg of melatonin 30 min before sleeping makes you feel 8-hours rested after 7-hours of sleep (but 9mg is harmful)
(2) that dosage has negligible cost
That's all you really needed to say.
It's jarring to me that you so meticulously analyze the cost of dosing with melatonin; once I know a cost is below some low threshold, I prefer not to think about it at all. I'd rather you took the same care into performing some objective tests of mental capability on varying amounts of sleep, so that it really means something when you say you gain an hour of wakefulness. Of course, I'd want this blinded as well, but I doubt you have convincing placebo pills available; besides, I don't mind taking something in hope of accruing some real and placebo benefits.
If your experience is typical, then the only reason people shouldn't be dosing melatonin is if there's some long-term health detriment (I don't have any mechanism in mind; it seems unlikely).
I was prescribed melatonin for a sleep complaint some years ago and noticed no detectable improvement: I didn't get to sleep more easily, wake up less during the night, or feel more refreshed in the morning. What might explain this?
Do you have a study that confirms your 'melatonin subtracts an hour' theory you could link to? My husband uses melatonin and can still easily spend 12 hours in bed. I've avoided using it, since I don't have difficulty actually falling asleep and I didn't want to sleep longer as a result of using it. You should probably argue that everyone should try using melatonin for a week or so, since the potential gains are large, not that everyone who doesn't use it is being foolish. The whole argument falls apart if your base assertion is wrong, and you provide no evidence that the effect melatonin has on you generalizes to everyone. That being said, I am glad you shared this information.
Since I wrote this post, the following comments have come in. If I missed any, please PM or reply with links. The following entries are compiled from LessWrong, #lesswrong, gwern.net, Reddit, or Google+, generally in relation to this melatonin essay; I classify as using/positive anyone who is using melatonin or will likely use melatonin in the future for any reason (eg. daily use counts, but so does using it for jetlag), and as negative/null anyone who is no longer using it for any reason (eg. found a better way to induce vivid dreams, or found it made the...
I'm interested in knowing how you came up with the conclusion that it reduces your sleep need (or bed time) by one hour?
I can understand that taking melatonin would reduce bed time if it made you fall asleep faster, but personally I've had no trouble falling asleep quickly so there would basically be no difference. (In fact, I've experienced an opposite effect with the couple of times I've taken melatonin.)
It takes a lot more than just casual observations to conclude that there are other effects. Basically you'd need to measure your sleep debt some way an...
I looked up melatonin in FASS. The only melatonin based drug that's legal in Sweden is called Cirkadin. It has 2mg of melatonin per pill. It is a prescription drug, and only recommended for patients who are older than 55.
There are some side effects. Common ones are headaches, inflammation of the throat (lasting 1-2 days, they don't say if that's alltogether or after you quit melatonin), back pain and weakness (asteni). That's probably the real cost of using it.
Melatonin induces CYP3a in vitro. If it does that in the blood as well it will decrease the effe...
This does not work with teenagers, in whom bedtimes are practically unenforceable, but whose need to get to school in a timely fashion does a reasonable job of imposing a getting-up time. They are chronically sleep deprived. Smaller children, I imagine, have even less of a chance of managing the feat.
Thanks for posting this. You've done a good job of explaining how melatonin is insanely cost-effective. I do want to mention some caveats for anyone who is interested in trying melatonin.
I use melatonin nightly. I admit it's effective, but it's no panacea. You can't drink a cup of coffee, then take some melatonin and expect to fall right asleep. It can also help to keep the lights low for about 30 minutes before bed time. Finally, one side-effect of melatonin is that I often have very vivid dreams. I don't know of any studies confirming this, but there are quite a few anecdotes.
I've been using melatonin for years now. I recommended it as part of my health article a while back.
When I started I was careful to always take a <1mg dose. More recently (due to laziness really) I have just been taking whatever dose I found at the store without bothering to break the tablets apart.
Your article reminded me that the current 3mg tabs I have are probably too much. This dose still works fine for me, but I have noticed a higher incidence of headaches which could be connected.
PS, good work on your site gwern, I find your research useful.
My rule of thumb is melatonin subtracts an hour. That is: if one slept for 7 hours, one awakes as refreshed as if one had slept for 8 hours (and so on).
That's the opposite of my impression. I used melatonin for several months because it made it so easy to get to sleep. But I found that I needed to get a full 8 hours of sleep to wake up refreshed. I attributed that to some vague notion that the melatonin needed that much time to work its way through my system. At any rate, waking up after less than 8 hours seemed harder than it does without melatonin...
Melatonin trip report:
I've used melatonin for two weeks now. Taking 1.5mg consistently puts me to sleep within an hour. Also gives me vivid dreams. Makes it easier to go to sleep at the same time every night. Does not make it easier or harder to get up in the morning.
Verdict: melatonin is worth it.
3. Melatonin supplements just doesn’t work on you, period.
• Possible, but unlikely. This isn’t some mental trick - it’s a fundamental fact of mammalian biology.
The presence of melatonin in the body and its function there may be fundamental, but that does not make the effect of supplements equally fundamental. The role of vitamins is also fundamental, but you yourself cite evidence that vitamin supplementation for all is a bad idea. I don't see anything in what you have written to suggest that melatonin supplementation for all should be any different.
FWI...
The gains I have laid out are significant enough I consider it irrational for someone not to use it, unless:
You forgot
(Pretend that 1 is a 9. It is. Honestly. It just shows up as a 1.)
Since the big criticism of this article was lack of proof of safety, which is a negative and rather hard to prove (who's going to pay for a long-term study of safety when there are lots of reasons to think there is no problem and bigger fish to fry?), I've given a try of the opposite - showing that melatonin supplementation improves health; see the last paragraph of http://www.gwern.net/Melatonin#use
If our computer use is massively cutting our melatonin secretion and lack of melatonin causes noticeable health problems in the most studied group (shift worke...
Do you have any evidence that melatonin is not addictive or habit forming? I have not read the studies you have linked; do they explicitly suggest this?
My understanding is that taking melatonin regularly will cause your body to naturally produce less melatonin, especially when taken in the 3mg dosage. Most doctors recommend a dosage of 1mg as that is closer to the body's natural amounts of melatonin.
Anyways, I think it's worth trying melatonin for one of the side effects, weird, vivid, possibly lucid dreams. Or possibly just increased dream recall.
I tried melatonin for the first time last night after reading this article. I will report back my results in two weeks.
The outcome was an abort for a reason that's not generally applicable. (Using melatonin prevented me from waking up in response to high or low blood sugar, which matters only for diabetics like me, but is an absolute showstopper, since failing to wake up in response to low blood sugar could be fatal.)
Miscellaneous note: Mentally, I find it easier to try to write for a non-LW audience, hence the tedious explanation & justification of the minimum wage assumption.
I have delayed-phase sleep disorder - I would say I "suffer" from it but it's really only a problem when a 3-10 sleep schedule is out of the question (as it is now, since I currently work 9-5). It's simply impossible for me to fall asleep before 2 or 3 am unless I am extremely tired. In addition, I'm a light sleeper, and have never been able to sleep while traveling or, in fact, whenever I'm not truly horizontal. I took melatonin to help with this for a couple years (at a recommended 0.3 mg dose), and it worked extremely well. However, I experien...
Consider me another data point for "found no obvious effect". I took a 3mg dose nightly for a couple weeks and didn't notice it doing much of anything. It might have made me feel a bit drowsy, but only briefly, and to a sufficiently small degree that I considered a biased perception due to wondering if I'd feel sleepier to be a more plausible hypothesis.
In either case, it certainly did not produce enough drowsiness to make "going to sleep" seem unusually appealing, and going to bed after taking it remained a conscious, deliberate exerci...
And what evidence do you have that it works the way you say without side effects... ? My self-experiments with Provigil say that nothing can reduce my long term need to sleep without severely deteriorating my mental capacity. Plural of anecdote being data...
I have a job that doesn't require me to be awake at any particular time, and my sleep schedule varies quite a bit (in the way that you'd expect; my average day is ~25 hours). I just purchased some melatonin tablets (this brand/dosage). Other than having heard it recommended as a sleep aid before, and having heard that it's not effective for everyone, I know nothing about it but what's in this post, which should minimize any psychosomatic effects. What's the most useful datapoint(s) that I could provide to the community?
Edit: Found a more accurate link to the pills I have.
I stopped taking melatonin because I'd wake up extremely groggy. Although I also have a (bad?) habit of covering my eyes with a pillow or the corner of a blanket AND I have a light blocking shade in my room, so I was basically waking up in complete darkness. One thing I considered to counteract this was plugging a lamp into a "vacation" timer and having it fire up around 7a or 8a or so. I think I'll stop at the hardware store and give the whole setup, melatonin and all, another try after reading this post.
I take it occasionally (mainly when I need to shift my sleep schedule back a few hours very quickly), but have kept from taking it nightly due to worries of tolerance and such.
Have you noticed any tolerance or know of any studies that cover this?
The metastudy I saw mentioned many studies finding no effect due to large doses causing tolerance effects (and recomended doses of no more than 1mg, IIRC).
I also never heard of the 'subtract an hour' part, nor have I noticed it on my own (though I use it in times when it'd be harder to notice). Has anyone else not...
I attempted to use melatonin on myself in a pseudo-study and wrote about it here: http://lesswrong.com/r/discussion/lw/ngj/open_thread_april_4_april_10_2016/d7s5
I'm a bit skeptical about this but it's an interesting idea and seems relevant to this discussion. It might even be a useful way of making any self-tests of melatonin slightly more scientific. I haven't tried it myself yet.
Sleep Cycle Analyzes Your Sleep Patterns for a Better Wake-Up
I've taken melatonin a few times-- it put me to sleep, though I think the onset of sleepiness was more like an hour after taking it. If I overrode the mild sleepiness, the melatonin's effect was lost,.
I've heard that people's natural production of melatonin declines with age, so there may be an age below which the pills don't make a huge difference for most people.
"It really amazes me that melatonin is available in any pharmacy," Bentley said. "It is a powerful hormone, and yet people don't realize that it's as 'powerful' as any steroid. I'm sure that many people who take it wouldn't take steroids so glibly. It could have a multitude of effects on the underlying physiology of an organism, but we know so little about how it interacts with other hormone systems."
Popular supplement melatonin found to have broader effects in brain than once thought (2005)
No. Your excerpt was a poor exposition of the standard precautionary principle I regard as entirely useless.
Reading your link now, I have even less reason to pay attention to it. It's a random press release about a presumably small unreplicated animal study in a species I don't know to be particularly germane to humans (eg. chimps) about changes of unclear importance in a body system with no human analogues ("In birds, switching off GnRH causes the gonads -- testes and ovary -- to shrink as part of the birds' yearly cycle.") with doses potentially high enough to be completely irrelevant to human supplementation (injecting melatonin?). I haven't even read the study!
Combine all the conditionals here (the smallness and lack of replication alone knocks down the chance this means anything about anything to well under 50%), and I don't see why I would update at all (not being an AI or anything which can represent degrees of belief with 64-bit floats).
If anything, I think this sort of study is a good example of why animal studies should be ignored in discussing supplements.
On the general topic of animal model external validity & translation to humans (with obvious relevance to supplements & nootropics), here are the major systematic reviews, meta-analyses, and articles criticizing the routine failures of animal models to provide any meaningful information about dangers or benefits in humans, and documenting the even lower quality of animal experiments than usual in (human) medicine or psychology:
I'm confused on why the value of an hour isn't just your hourly wage. If you value your time more, you'd be working less. If you value your time less, you'd be working more. Since you clearly spend exactly as much time working as you do, you value your free time that much.
Then again, that only works if you value the time you're trying to fall asleep or are asleep the same as you value work.
Back on topic: I've heard that taking melatonin makes your body produce less of it, though doing so the the point of having less total melatonin seems unlikely, and even if it's only a slight increase, it's only a few cents. Also, I didn't exactly hear it from a reliable source.
On the general topic of animal model external validity & translation to humans (with obvious relevance to supplements & nootropics), here are the major systematic reviews, meta-analyses, and articles criticizing the routine failures of animal models to provide any meaningful information about dangers or benefits in humans, and documenting the even lower quality of animal experiments than usual in (human) medicine or psychology:
I spent the day reading up on the topic. The transfer or replication rates range from around 40% on the upper end to zero in some cases involving as many as hundreds of attempts to transfer. The methodological quality of the animal studies are usually terrible with hardly any blinding and randomization rare (and when it is done, one author indicates that researchers surveyed would say it was done as informally as grabbing random mice/rats out of the cages), and the publication biases at play seem to be even larger than in human studies.
(Reading the papers, I found myself disturbed by the ethical implications: when human studies fall to publication bias, 'all' that does is waste many millions of dollars and impede the progress of science and substantially inconvenience many subjects and put them at risk; but when at least a third of all animal experiments never get written up, and half of chimpanzee medical/biological studies never get cited when they do get published, that means that the billions of animals gone through every few years are mutilated and tortured and killed for absolutely nothing at all. This makes me much more sympathetic to the NIH's recent retirement of its research chimpanzees.)
There are some interesting examples, though; from one of the links:
From the toxicology paper:
From my appendix:
(I imagine that the latter is particularly relevant; how many thousands of mouse studies on inflammation were part of the evidence base for those <150 clinical trials? Probably quite a few. And it seems that they all are essentially irrelevant to anything in humans. Now, imagine the translation rate for bird to primate, based on a bird system which doesn't even exist in humans...)
Thanks for doing all this research! You should make a page on gwern.net, you're wasting your talents going in to this kind of depth in an ancient comments thread ;)