This is why taking melatonin tablets before is recommended by gwern and others. I second this recommendation. Though not FDA-approved, there seem to be little in the way of negative side effects and they make it much easier to fall asleep.
(This paragraph is retracted. See jacob_cannell’s post below.) Melatonin is marginally useful on occasion, but overrated. People who take melatonin for a long period of time likely don’t get much benefit from it. The longest duration study I’ve seen on melatonin suggested the benefits did not last a year for over 90% of participants. To be fair, the study participants had delayed sleep phase disorder (DSPD), so perhaps the response is different for normal people. But I consider that unlikely given that both normals and people with DSPD take melatonin to shift their circadian rhythm and fall asleep at a more appropriate time.
My opinion is that sleep hygiene is far more useful than any drug or supplement. I’m a particular fan of the so-called “Bootzin technique”, which basically makes falling asleep quickly at an appropriate time and place a habit. Also, if you suspect you might have a sleep disorder like sleep apnea or narcolepsy, it would be best to get that checked out.
The longest duration study I’ve seen on melatonin suggested the benefits did not last a year for over 90% of participants.
This sounded interesting to me so I followed your link to the study. I think you may have seriously misinterpreted it. They treated the subjects with 5mg nightly melatonin for 6 weeks. After the treatment ended, they followed up over a year later. 91.5% reported a relapse to previous sleeping patterns after stopping melatonin treatment.
I’m not sure how you could conclude “people who take melatonin for a long period of time don’t get much benefit from it” based on that study.
You are absolutely correct. Thank you. It appears I misread the abstract, which is worrisome, but I’m glad to have been corrected. I added a note to the original post retracting that statement. I also tracked down another post where I made the same claim and retracted the claim there. I’ll have to read things more carefully*.
I am now not aware of any studies which looked at the long term effects of melatonin supplementation. I would not be surprised at all if oral melatonin lost effectiveness over time, though I recall gwern saying he did not see evidence of a tolerance effect. I also think people claim many positive things about melatonin that don’t seem to check out (negative things, too, in my case). In particular, I don’t see solid reasons to believe melatonin improves sleep quality. It seems to basically only shift the circadian cycle, which can be useful in circadian rhythm disorders and insomnia. I don’t have my papers available right now, but I can pull up a few on melatonin’s effect on sleep architecture if you are interested. (I’ll probably do this anyway just to make sure I read them correctly after the mistake you highlighted!)
* To be honest, when I was reading studies about sleep a year-ish ago, I was very tired. My motivation was understanding why I might be so tired during the day. I should not be surprised if I was sloppy through this process. I decided I’m a long sleeper, and I’ve been considerably better since trying to get as much a sleep as possible.
I’ve been taking melatonin for about … 5 years, so if you DO have some evidence for any long term effects, that is of course something I’m interested in.
I suspect that there is some genetic variation in the circadian system such that melatonin isn’t one size fits all. In particular I have a pretty long sleep onset. Before melatonin, my estimate of my own sleep estimate was about 30 minutes to an hour. With melatonin it’s usually less than 30 minutes, although I’ve adjusted my dose down to 2 mg or so trying to get the minimum that works.
For me the benefit seems to be more than the 30 minutes saved, my sleep also becomes somewhat more effective.
(This paragraph is retracted. See jacob_cannell’s post below.) Melatonin is marginally useful on occasion, but overrated. People who take melatonin for a long period of time likely don’t get much benefit from it. The longest duration study I’ve seen on melatonin suggested the benefits did not last a year for over 90% of participants. To be fair, the study participants had delayed sleep phase disorder (DSPD), so perhaps the response is different for normal people. But I consider that unlikely given that both normals and people with DSPD take melatonin to shift their circadian rhythm and fall asleep at a more appropriate time.
My opinion is that sleep hygiene is far more useful than any drug or supplement. I’m a particular fan of the so-called “Bootzin technique”, which basically makes falling asleep quickly at an appropriate time and place a habit. Also, if you suspect you might have a sleep disorder like sleep apnea or narcolepsy, it would be best to get that checked out.
Also, while we’re on the topic of sleep control, there are a number of mathematical models of sleep which you might find useful. I have no idea how accurate these models are, or if they offer any hints about controlling sleep beyond what’s obvious. Here’s a review paper about what seems to be the most popular model. SuperMemo, the spaced repetition memory software, has a module which uses something like this model.
This sounded interesting to me so I followed your link to the study. I think you may have seriously misinterpreted it. They treated the subjects with 5mg nightly melatonin for 6 weeks. After the treatment ended, they followed up over a year later. 91.5% reported a relapse to previous sleeping patterns after stopping melatonin treatment.
I’m not sure how you could conclude “people who take melatonin for a long period of time don’t get much benefit from it” based on that study.
You are absolutely correct. Thank you. It appears I misread the abstract, which is worrisome, but I’m glad to have been corrected. I added a note to the original post retracting that statement. I also tracked down another post where I made the same claim and retracted the claim there. I’ll have to read things more carefully*.
I am now not aware of any studies which looked at the long term effects of melatonin supplementation. I would not be surprised at all if oral melatonin lost effectiveness over time, though I recall gwern saying he did not see evidence of a tolerance effect. I also think people claim many positive things about melatonin that don’t seem to check out (negative things, too, in my case). In particular, I don’t see solid reasons to believe melatonin improves sleep quality. It seems to basically only shift the circadian cycle, which can be useful in circadian rhythm disorders and insomnia. I don’t have my papers available right now, but I can pull up a few on melatonin’s effect on sleep architecture if you are interested. (I’ll probably do this anyway just to make sure I read them correctly after the mistake you highlighted!)
* To be honest, when I was reading studies about sleep a year-ish ago, I was very tired. My motivation was understanding why I might be so tired during the day. I should not be surprised if I was sloppy through this process. I decided I’m a long sleeper, and I’ve been considerably better since trying to get as much a sleep as possible.
I’ve been taking melatonin for about … 5 years, so if you DO have some evidence for any long term effects, that is of course something I’m interested in.
I suspect that there is some genetic variation in the circadian system such that melatonin isn’t one size fits all. In particular I have a pretty long sleep onset. Before melatonin, my estimate of my own sleep estimate was about 30 minutes to an hour. With melatonin it’s usually less than 30 minutes, although I’ve adjusted my dose down to 2 mg or so trying to get the minimum that works.
For me the benefit seems to be more than the 30 minutes saved, my sleep also becomes somewhat more effective.