Scientific Opinion on the substantiation of a health claim related to melatonin and reduction of sleep onset latency (ID 1698, 1780, 4080)
pursuant to Article 13(1) of Regulation (EC) No 1924/2006[sup]1[/sup]
EFSA Panel on Dietetic Products, Nutrition and Allergies (NDA)2, 3
European Food Safety Authority (EFSA), Parma, Italy
Słowa kluczowe:
Melatonin
health claims
insomnia
sleep onset latency
3. Naukowe uzasadnienia wpływu na zdrowie człowieka - Zmniejszenie trudności w zasypianiu
-
Melatonina pomaga skrócić czas potrzebny na zaśnięcie
A claim on melatonin and reduction of sleep onset latency and improvement of sleep quality has already been assessed with an unfavourable outcome (EFSA Panel on Dietetic Products Nutrition and Allergies (NDA), 2010), based on a meta-analysis (Buscemi et al., 2006) of randomised controlled trials (9 studies) conducted in subjects with sleep disorders under sleep restrictions (subjects with changes in sleep pattern, e.g. after jet lag, shift work or induced insomnia), which did not show an effect of melatonin consumption on sleep onset latency or sleep quality.
The references provided for the scientific substantiation of the present claim included textbooks, an extract from the proceeding of a colloquium, and narrative reviews, which did not provide any original data for the scientific substantiation of the claim. An abstract and a reference were also submitted in
which the information provided regarding the study design, methodology and statistical analyses was insufficient for a full scientific evaluation. Some of the references were on substances other than melatonin (lutein and zeaxanthin), on intranasal administration of melatonin, which is not a route considered relevant for human nutrition, or dealt with outcomes, such as subjective feeling of jet lag or subjective sleepiness (without any specific measurements related to sleep onset latency), chemistry of ligands at the membrane receptors of melatonin, melatonin delivery in pharmacological formulations, the secretion of melatonin, and modification of the biological rhythms of core body temperature, cortisol and melatonin secretion in response to nocturnal bright light stimuli in combination with melatonin supplementation, which were unrelated to the claimed effect. Two references reported on the effects of melatonin either in healthy subjects after an overnight sleep deprivation (Tzischinsky and Lavie, 1994) (sleep restriction has already been evaluated in the previous opinion on melatonin), or in subjects with secondary sleep disorders (women with mild or moderate asthma) (Campos et al., 2004). The Panel considers that no conclusions can be drawn from these references for the scientific substantiation of the claim.
Three meta-analyses of controlled trials, which assessed the effect of melatonin consumption on sleep onset latency in healthy subjects without insomnia (Buscemi et al., 2004), in subjects with primary sleep disorders (Buscemi et al., 2005), or in healthy subjects with or without insomnia (Brzezinski et al., 2005), were not considered in the aforementioned published opinion (EFSA Panel on Dietetic Products Nutrition and Allergies (NDA), 2010), and included all the individual human intervention studies provided for the scientific substantiation of the present claim.
The meta-analysis by Buscemi et al. (2004) considered controlled trials (published between 1966 and 2003) which had investigated the effect of melatonin administration on objective sleep outcome measures, including sleep onset latency, in adult normal sleepers. Sleep outcomes were measured by polysomnography, actigraphy or questionnaires. Twenty studies which compared the effect of melatonin (n=225 subjects) vs. placebo (n=206 subjects) met the inclusion criteria. Eighteen studies had a cross-over design. Study duration ranged from one night to five weeks. The doses of melatonin administered ranged from 0.1 to 100 mg, and were categorised according to the following levels: <1 mg; 1-3 mg; 4-5 mg; 6-10 mg; >10 mg. The meta-analysis was performed using a random effects model, and fixed effects were considered in a sensitivity analysis. Study quality was low-to-moderate (i.e. none of the studies reported a power calculation for the primary outcome, and although all studies involved a placebo control, many did not apply random allocation to the intervention group). The meta-analysis showed a statistically significant decrease in sleep onset latency following melatonin administration compared to placebo (combined weighted mean difference (WMD): -3.9 min; 95% CI -5.3, -2.6). The sub-group analysis regarding the dosage of melatonin showed a statistically significant decrease in sleep onset latency after melatonin administration compared to placebo for all considered dosage ranges (<1 mg, five studies; 1-3 mg, 10 studies; 4-5 mg, six studies; 6-10 mg, seven studies) except for the highest doses (>10 mg, two studies).
The meta-analysis by Buscemi et al. (2005) considered controlled studies (published between 1966 and 2004) which had investigated the effect of melatonin administration on objective sleep outcome measures, including sleep onset latency, in subjects with sleep disorders (insomnia, delayed sleep-phase syndrome). Sleep outcomes were measured by polysomnography, actigraphy or questionnaires. Fourteen studies, which compared the effect of melatonin administered before bedtime (n=218 subjects) vs. placebo (n=207 subjects), met the inclusion criteria. Eleven studies had a cross-over design. The studies were conducted in children (2 studies), adults aged 19-65 (7 studies), and adults aged 66 years and older (5 studies). The meta-analysis was performed using a random effects model, owing to the high heterogeneity observed in the study results. Study quality was ranked as moderate to high by the authors. The meta-analysis showed a statistically significant decrease in sleep onset latency after melatonin administration compared to placebo (WMD: -11.7 min; 95 % CI -18.2, -5.2). The sub-group analysis regarding the dosage of melatonin showed a statistically significant decrease in sleep onset latency after melatonin administration compared to placebo for the dosage ranges 1-3 mg (six studies), but not for <1 mg (two studies) or for 4-5 mg (seven studies).
The Panel notes that, in the two meta-analyses (Buscemi et al., 2004; Buscemi et al., 2005), the lower end of the dose ranges of the sub-group analyses for which a significant effect of melatonin on sleep onset latency was consistently found is 1 mg.
The meta-analysis by Brzezinski et al. (2005) included double-blind randomised controlled trials (published between 1980 and 2003) which investigated the effect of melatonin consumption on objective sleep outcome measures. Sleep outcomes were measured by polysomnography, actigraphy or the index finger switch depression method. Twelve of the 17 studies included in the meta-analysis provided measures of sleep onset latency. Six studies were conducted on healthy volunteers, four on subjects with insomnia, one on both healthy volunteers and insomniacs, and one on a combination of institutionalised and free-living subjects with insomnia. Sample size ranged from 6 to 30 subjects, all studies had a cross-over design with or without a wash-out phase, duration ranged from one night to three weeks, melatonin doses from 0.1 to 40 mg/day, and either one or multiple melatonin doses were tested in the same study. Considering these 12 studies (n=172 subjects), melatonin significantly reduced sleep onset latency by 3.9 min (95 % CI -2.5, -5.4) compared to placebo. Removing the two smallest studies (6 and 8 subjects, respectively) from the analysis, because of the presence of outliers in the data set or heterogeneity of the data, led to a reduction of sleep onset latency of 7.4 min (95 % CI -5.1, -9.8) in favour of melatonin. The Panel considers that the results of this meta-analysis can be used as evidence for the substantiation of the claim, but notes some methodological limitations, including the lack of extensive sensitivity analysis or adjustment for the different doses used in the various studies.
The Panel notes that the major difference between the four meta-analyses (Brzezinski et al., 2005; Buscemi et al., 2004; Buscemi et al., 2005; Buscemi et al., 2006) was the population groups considered. The Buscemi et al. (2006) analysis considered in the previously published opinion (EFSA Panel on Dietetic Products Nutrition and Allergies (NDA), 2010) did not show a significant effect of melatonin consumption on sleep onset latency in a sub-group analysis considering subjects with sleep restrictions (after jet lag, shift work or induced insomnia). In contrast, the analysis by Buscemi et al. (2004) in normal sleepers, together with the sub-analysis by Brzezinski et al. (2005) in healthy subjects or subjects with no relevant medical condition other than insomnia (excluding studies on sleep restriction), and the analysis by Buscemi et al. (2005) in subjects with primary sleep disorders (insomnia, delayed sleep-phase syndrome) found statistically significant effects of melatonin consumption on sleep onset latency. Overall, the Panel notes that the three meta-analyses considered in this opinion showed a significant effect of melatonin on sleep onset latency. In the two meta-analyses (Buscemi et al., 2004; Buscemi et al., 2005), the lower end of the dose ranges of the sub-group analyses for which a significant effect of melatonin on sleep onset latency was consistently found is 1 mg.
In weighing the evidence, the Panel took into account that a meta-analysis of controlled human intervention studies in normal sleepers indicated a statistically significant reduction of sleep onset latency following melatonin consumption, and that these results were supported by two meta-analyses of controlled human intervention studies, one in subjects with primary sleep disorders and one in healthy subjects combined with subjects with insomnia.
The Panel concludes that a cause and effect relationship has been established between the consumption of melatonin and reduction of sleep onset latency.
