Nightly treatment of primary insomnia with prolonged release melatonin for 6 months: a randomized placebo controlled trial on age and endogenous melatonin as predictors of efficacy and safety

Abstract

Background: Melatonin is extensively used in the USA in a non-regulated manner for sleep disorders. Prolonged release melatonin (PRM) is licensed in Europe and other countries for the short term treatment of primary insomnia in patients aged 55 years and over. However, a clear definition of the target patient population and well-controlled studies of long-term efficacy and safety are lacking. It is known that melatonin production declines with age. Some young insomnia patients also may have low melatonin levels. The study investigated whether older age or low melatonin excretion is a better predictor of response to PRM, whether the efficacy observed in short-term studies is sustained during continued treatment and the long term safety of such treatment.

Methods: Adult outpatients (791, aged 18-80 years) with primary insomnia, were treated with placebo (2 weeks) and then randomized, double-blind to 3 weeks with PRM or placebo nightly. PRM patients continued whereas placebo completers were re-randomized 1:1 to PRM or placebo for 26 weeks with 2 weeks of single-blind placebo run-out. Main outcome measures were sleep latency derived from a sleep diary, Pittsburgh Sleep Quality Index (PSQI), Quality of Life (World Health Organzaton-5) Clinical Global Impression of Improvement (CGI-I) and adverse effects and vital signs recorded at each visit.

Results: On the primary efficacy variable, sleep latency, the effects of PRM (3 weeks) in patients with low endogenous melatonin (6-sulphatoxymelatonin [6-SMT] <or=8 microg/night) regardless of age did not differ from the placebo, whereas PRM significantly reduced sleep latency compared to the placebo in elderly patients regardless of melatonin levels (-19.1 versus -1.7 min; P = 0.002). The effects on sleep latency and additional sleep and daytime parameters that improved with PRM were maintained or enhanced over the 6-month period with no signs of tolerance. Most adverse events were mild in severity with no clinically relevant differences between PRM and placebo for any safety outcome.

Conclusions: The results demonstrate short- and long-term efficacy and safety of PRM in elderly insomnia patients. Low melatonin production regardless of age is not useful in predicting responses to melatonin therapy in insomnia. The age cut-off for response warrants further investigation.

Trial registration: ClinicalTrials.gov NCT00397189.

Figure 1

Overall study patient disposition (CONSORT).

Figure 2

The effect of 3 weeks treatment with prolonged release melatonin (PRM) and placebo on sleep latency in the low excretors and age 65-80 populations. Mean+standard error of mean values of the change from baseline in sleep latency from the sleep diary following 3 weeks of double blind treatment of low excretors (N = 86 per group) and age 65-80 years (N = 137 PRM, 144 placebo) populations with PRM and placebo. Asterisks denote significant difference between PRM and placebo groups (**P < 0.01).

Figure 3

Sleep latency during the treatment period. Mixed Effect Model for Repeated Measures (MMRM) predicted mean values (mean + standard error of mean) for sleep latency from the sleep diary, at baseline and weeks 3-29 of the double blind treatment periods, in those aged 65-80 years. Asterisks denote significant difference between prolonged release melatonin and placebo groups (*P < 0.05 ** P < 0.01). Numbers of patients analysed in each treatment time point are depicted

Figure 4

Time of going to bed during the treatment period. MMRM-predicted mean values (mean + standard error of mean) for time going to bed (hours relative to midnight) from the sleep diary at baseline and weeks 3-29 of the double blind treatment periods, Intention to Treat (ITT) 65-80 population. Asterisks denote significant difference between prolonged release melatonin and placebo groups (*P < 0.05 ** P < 0.01). Numbers of patients analysed in each treatment time point are depicted.