Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Melatonin for Sleep Disorders

Searching
Primary sleep disorder [1]
Secondary sleep disorder [2]
Sleep restriction [2]
Comment

Bandolier 82 looked at a good, large, randomised trial of melatonin for jet lag and found it not to be effective for sleep problems. Of course, melatonin may still have been useful for other sleep disturbances, though one would have had to be an enthusiast to bet on it. We now have two systematic reviews [1,2] to demonstrate that being a bit suspicious was probably the correct approach.

Searching


These linked reviews had a heroic searching strategy, covering electronic searching of more databases than most of us knew existed, plus bibliographies and some hand searching. The latest searches were in early 2004. For inclusion, studies had to be randomised trials involving humans with a sleep disorder (primary, secondary, or sleep disorder accompanying sleep restriction), compared melatonin with placebo, and reported one or more of a variety of sleep outcomes. Primary sleep disorders were those for which there was no medical condition (schizophrenia, developmental disability, depression) causing it. Studies reporting only adverse events were also included for safety analysis.

Primary sleep disorder [1]


Fourteen trials were included for efficacy and 10 for safety. Trial size ranged from 16 to 62 patients. The maximum number of patients upon which efficacy was based was 425, and for safety 281.

There was no difference between melatonin and placebo for all efficacy outcomes except sleep onset latency, which was 12 minutes less with melatonin, with a 95% confidence interval of 5 to 18 minutes. The 10 studies of better reporting quality and less likelihood of being biased were not significantly different for sleep latency. There was no difference in adverse event rates for headache, dizziness, nausea, or drowsiness.

Secondary sleep disorder [2]


For sleep disorder secondary to another medical condition, nine trials were included for efficacy and seven for safety. Trial size ranged from six to 157 patients. The maximum number of patients upon which efficacy was based was 382, and for safety 253.

There was no difference between melatonin and placebo for sleep onset latency, wakefulness after sleep onset, or minutes of restless eye movement sleep. There was a small (16 minute) significant increase in total sleep time. There was no difference in adverse event rates for headache, dizziness, nausea, or drowsiness.

Sleep restriction [2]


For sleep restriction studies, nine trials were included for efficacy and 10 for safety. Trial size ranged from 17 to 320 patients. The maximum number of patients upon which efficacy was based was 508, and for safety 560.

There was no difference between melatonin and placebo for sleep onset latency, wakefulness after sleep onset, or minutes of restless eye movement sleep. There was a small (18 minute) increase in total sleep time, based on a small number of patients. There was no difference in adverse event rates for headache, dizziness, nausea, or drowsiness.

Comment


Different doses of melatonin were given to different patients for different periods of time, and while similar outcomes were reported, this was not consistent throughout the studies. Most trials were trivially small. There was no obvious dose response. Any benefits for melatonin were small (Table 1), and derived from multiple comparisons. What is the likelihood of melatonin being a whizzo medicine for sleep disorders? Not likely, is it?


Table 1: Summary of results of trials of melatonin for sleep disorders



Number of patients
Patient group
Melatonin
Placebo
Conclusion
Primary sleep disorder
218
207
No effect on any outcome using better quality studies
Secondary sleep disorder
220
162
Increased sleep time by about 16 minutes
Sleep restriction
309
199
Increased sleep time by about 18 minutes, based on 200 patients only



In each review subgroup analysis with tiny numbers makes one worry. Is melatonin safe for short term use? Using the rule of three, we can only conclude that since no disaster has been seen in about 700 patients taking melatonin, we can be 95% sure that disasters will not occur more frequently than once in every 200 patients taking melatonin.

There is a call for large-scale randomised trials to prove the lack of efficacy of melatonin conclusively. Given the costs and difficulties of clinical research, and given that there is no whisker, no sniff of any useful efficacy, and given the unknown quantity of unpredictable but possibly harmful effects, is it ethically sound or reasonable for more trials, at least in these conditions?

References:



  1. N Buscemi et al. The efficacy and safety of exogenous melatonin for primary sleep disorders: A meta-analysis. Journal of General Internal Medicine 2005 20: 1151-1158.
  2. N Buscemi et al. Efficacy and safety of exogenous melatonin for secondary sleep disorders and sleep disorders accompanying sleep restriction: meta-analysis. BMJ 2006 doi:10.1136/bmj.38731.532766.F6 (published 10 February 2006).

previous or next story