Open Access Journal of Clinical Trials Dovepress

open access to scientific and medical research

Open Access Full Text Article ME T H OD OLOGy

Design and rationale of a randomized controlled

trial of melatonin supplementation in men
and women with the metabolic syndrome
This article was published in the following Dove Press journal:
Open Access Journal of Clinical Trials
5 March 2013
Number of times this article has been viewed

Paul D Terry 1 Background: The metabolic syndrome is a constellation of interrelated metabolic risk factors
Abhinav Goyal 2,3 that appear to increase the risk of atherosclerotic cardiovascular disease, type 2 diabetes mellitus,
Lawrence S Phillips 3 and possibly some cancers. Animal studies and observational clinical data in humans suggest that
Hillary M Superak 4 supplemental melatonin may ameliorate a number of components of the metabolic syndrome,
Michael H Kutner 4 including elevated glucose, elevated blood pressure, dyslipidemia, and obesity. The primary
objective of this clinical trial was to determine the feasibility, efficacy, and safety of melatonin
1
Departments of Public Health and
Surgery, University of Tennessee, supplementation in men and women with the metabolic syndrome.
Knoxville, TN, 2Department of Methods: Thirty-nine men and women of mixed race/ethnicity were enrolled into a randomized,
Epidemiology, Emory Rollins School double-blind, placebo-controlled clinical trial with two arms: placebo for 10 weeks followed by
of Public Health, 3Department of
Medicine, Emory School of Medicine, melatonin for 10 weeks, or vice versa, with an interval 6-week washout period, in a crossover
4
Department of Biostatistics and trial design. Outcome measures include metabolic syndrome components (blood pressure,
Bioinformatics, Emory Rollins School
glucose, triglycerides, high-density lipoprotein cholesterol, waist circumference), oxidative
of Public Health, Atlanta, GA, USA
stress, and inflammation biomarkers. These biomarkers, along with sleep duration and quality
and pretreatment endogenous melatonin levels, were measured to explore possible underlying
biologic mechanisms.
Discussion: This trial will provide knowledge of the effects of melatonin in metabolic syn-
drome subjects, and lay the groundwork for future clinical trials of melatonin in metabolic
syndrome subjects.
Keywords: melatonin, metabolic syndrome, diabetes, blood pressure, sleep

Introduction
The metabolic syndrome is a constellation of interrelated metabolic risk factors that
appear to increase the risk of atherosclerotic cardiovascular disease, type 2 diabetes
mellitus, and possibly some cancers.1–3 Although pharmacologic treatments for some
of the metabolic syndrome components exist, these treatments are not always effective
and are often associated with adverse effects. Therefore, there is considerable interest
in finding alternative treatments for the metabolic syndrome.
Melatonin (N-acetyl-5-methoxytryptamine) is an endogenous indoleamine com-
pound (chemical formula C H N O ). Melatonin has chronobiotic, antioxidant, anti-
13 16 2 2

Correspondence: Michael H Kutner inflammatory, and soporific properties. A major site of melatonin biosynthesis is the
Department of Biostatistics pineal gland, from which melatonin is released into the circulation in response to photic
and Bioinformatics, Emory Rollins School
of Public Health, 1518 Clifton Road NE, light/dark information transmitted from the retina to the suprachiasmatic nucleus of the
Atlanta, GA 30322-4201, USA hypothalamus.4 Pineal melatonin production and release follows a circadian pattern,
Tel 1 404 712 9708
Fax 1 404 727 1370
with increased release in darkness and decreased release during exposure to light. It
Email [email protected] is understood that one of the main physiologic functions of melatonin is to transmit

submit your manuscript | www.dovepress.com

Open Access Journal of Clinical Trials 2013:5 51–59 51
Dovepress © 2013 Terry et al, publisher and licensee Dove Medical Press Ltd. This is an Open Access article
http://dx.doi.org/10.2147/OAJCT.S39551 which permits unrestricted noncommercial use, provided the original work is properly cited.
Terry et al Dovepress

information about the light/dark cycle to the body, thereby crossover design, each subject serves as his or her own
allowing the body’s physiologic systems to orient themselves control. Recruitment, randomization, and subject follow-up
according to seasonal and circadian patterns.4 have been completed (Figure 1), and laboratory assays and
Animal studies and observational clinical data in humans data cleaning are nearly complete.
suggest that supplemental melatonin may ameliorate a Subjects in the MetSyn study were recruited from several
number of metabolic syndrome components, including sources: the population of an already completed study funded
elevated glucose, elevated blood pressure, dyslipidemia, by the National Institute of Diabetes and Digestive and Kidney
and obesity.5–9 The evidence also suggests that individu- Diseases of the National Institutes of Health, ie, Screening
als with metabolic syndrome components have suppressed for IGT: Glucose Challenge versus Predictive Model (SIGT,
melatonin levels, and hence may be particularly likely to R18DK066204; Lawrence S Phillips, principal investigator);
benefit from exogenous melatonin supplementation.10 On the Preventive Cardiology Clinic at the Emory Clinic in
the whole, evidence from animal experiments and obser- Atlanta, GA, USA; and advertisements placed on the campus
vational epidemiologic studies suggest that endogenous of the Emory Clinic. The purpose of the SIGT study was to
melatonin or melatonin supplementation may protect against test the hypothesis that screening by a one-hour oral glucose
elevated glucose/insulin resistance, elevated blood pressure, challenge test would have good predictive ability to identify
dyslipidemia, and perhaps abdominal obesity. However, the impaired glucose tolerance, and would be superior to con-
evidence is insufficient in several ways, including the lack ventional predictive models in both diagnostic efficiency and
of randomized clinical trials in humans, particularly among cost-effectiveness. SIGT enrolled a sample of 525 metabolic
individuals with the metabolic syndrome.9 syndrome subjects who were Emory University employees.
The purpose of this paper is to present the design, The MetSyn study enrolled a total of 39 men and women
rationale, and methodologic considerations of a randomized, (aged 30–79 years) who met the criteria for metabolic syn-
double-blinded, placebo-controlled, clinical trial of melatonin drome (described below). Subject visits for the clinical trial
supplementation in men and women with the metabolic were at the Emory University General Clinical Research
syndrome. This trial will assess the feasibility of recruitment Center, also known as the Atlanta Clinical and Translational
of metabolic syndrome subjects and will determine if treatment Science Institute. Subjects were compensated financially for
with melatonin (10 weeks) can ameliorate one or more of the their participation in the trial.
metabolic syndrome components, ie, blood pressure, glucose,
triglycerides, high-density lipoprotein (HDL) cholesterol, and Inclusion and exclusion criteria
waist circumference. Our long-term goal is to determine the Consistent with the criteria established by the National Heart,
safety and efficacy of melatonin to reduce the risk of develop- Lung, and Blood Institute and the American Heart Asso-
ment of type 2 diabetes and cardiovascular disease in high-risk ciation,1 an individual was defined as having the metabolic
individuals with the metabolic syndrome. syndrome if he/she had three or more of the following five
conditions: elevated fasting glucose ($100 mg/dL), enlarged
Materials and methods blood pressure (.130 mmHg systolic or .85 mmHg dia-
Overview stolic), elevated triglycerides ($150 mg/dL), reduced
The Melatonin and the Metabolic Syndrome (MetSyn) study
HDL cholesterol (,40 mg/dL in men, ,50 mg/dL in
was a randomized, double-blind, placebo-controlled clini-
cal trial with two arms, ie, placebo for 10 weeks followed women), and enlarged waist circumference ($102 cm in
by melatonin for 10 weeks or vice versa, with an interval men, $88 cm in women). Subjects previously diagnosed
6-week washout period, in a crossover trial design.11 Thirty- with diabetes mellitus were excluded. To avoid the
nine subjects were randomized to one of the following two possibility of an increased risk of developmental
treatments, ie, placebo once daily one hour before bedtime disorders, we excluded women who were pregnant, breast-
for 10 weeks or 8.0 mg oral melatonin supplementation feeding, or attempting or planning to attempt conception.
once daily one hour before bedtime for 10 weeks. At the We also excluded men and women with the fol- lowing
end of the first 10-week period, each subject underwent a medical conditions: active malignancy (other than non-
6-week washout period in which no pills were taken. At the melanoma skin cancer); hyperthyroidism or hypothyroidism;
end of the washout period, the subjects were then crossed heart failure (New York Heart Association functional class
over to the alternative treatment for 10 weeks. Using this III or IV); history of myocardial infarction, bypass surgery,
angioplasty, or stroke within the past year; active liver disease;

52 submit your manuscript | www.dovepress.com

Open Access Journal of Clinical Trials 2013:5
Dovepress
 Dovepress Melatonin supplementation and metabolic syndrome
active pancreatitis or pancreatic insufficiency; active
peptic

Open Access Journal of Clinical Trials 2013:5 submit your manuscript | www.dovepress.com
53
Dovepress
Terry et al Dovepress

Assessed for eligibility

(n = 72)

Excluded (n = 33)
 Metabolic syndrome criteria not met (n = 17)
 Noncompliant during run-in (n = 2)
 Diabetes mellitus (n = 4)
 Unable to schedule (n = 4)
 Calcium channel blockers (n = 3)
 Active malignancy (n = 1)
 HIV-positive (n = 1)
 Taking sleep aids (n = 1)
Randomized
(n = 39)

Placebo first Melatonin first

(n = 20) (n = 19)

Lost to follow-up (n = 2)
 Subject withdrew upon physician’s
advice (n = 1)
Lost to follow-up
 Subject completed 10-week melatonin
(n = 0) treatment, but self-withdrew during
the subsequent 10-week placebo
treatment (n = 1)

Number to be analyzed Number to be analyzed

(n = 20) (n = 19)

Figure 1 Subject recruitment flow chart.

ulcer disease; previous gastrectomy or bowel resection; intesti- their physicians, and expressed interest in the study. After
nal malabsorption syndrome; narcotic or alcohol dependence; informed consent, we confirmed further eligibility after
recent significant ($10%) intentional or unintentional weight reviewing medical charts, and conducting in-person inter-
loss or weight gain; end-stage kidney disease on dialysis; views, a medical examination, and screening blood tests. If
taking immunosuppressive therapy (eg, systemic corticoster- the subject was eligible after all these steps and still willing
oids, azathioprine, methotrexate, cyclophosphamide); taking to participate, the subject underwent a 10-day placebo run-in
lithium therapy; or any condition with a life expectancy of during which time adherence to therapy was assessed. Even
less than 2 years. We also excluded subjects taking calcium though only placebo was administered during the run-in
channel blockers because a clinical study of melatonin and phase, subjects were informed only that they were taking
blood pressure suggested that melatonin supplementation may “study drug.” Subjects who had taken at least 80% of their
have increased blood pressure in persons taking calcium chan- tablets (placebo) without bothersome side effects were
nel blockers.12 Finally, we limited enrollment to nonsmokers randomized.
based on evidence of potential confounding from tobacco
smoking.10 Randomization
Randomization was performed independently by a bio-
Recruitment and screening statistician not involved in the study. Four randomly
We screened men and women who appeared to meet eli- generated permuted blocks of size 10 (five placebo and
gibility criteria, received support for participation from five melatonin) were generated. The order of randomization

54 submit your manuscript | www.dovepress.com

Open Access Journal of Clinical Trials 2013:5
Dovepress
 Dovepress Melatonin supplementation and metabolic syndrome

for the estimated 40 subjects was placed in sealed enve- the study. The project manager gave the bottles of pills to
lopes and given to the clinical pharmacy department. The the subjects at each visit.
pharmacy used these envelopes to determine the random- Random samples of the melatonin lot were sent to
ization status for each subject in the study. Subjects were consumerlab.com for independent verification. Consumerlab.
initially randomized to one of the following two treatments: com uses high-performance liquid chromatography and
placebo once daily one hour before bedtime for 10 weeks inductively coupled plasma mass spectrometry analytic
or 8.0 mg melatonin once daily one hour before bedtime methods to verify the biochemical composition of various
for 10 weeks. At the end of the first 10-week period, each drugs/supplements, and has extensive experience testing
subject underwent a 6-week washout period in which no melatonin preparations.
pills were taken. At the end of the washout period, the In this trial, subjects were instructed to take melatonin one
subject was crossed over to the alternative treatment for hour before bedtime, typically close to 9 pm. The rationale
10 weeks (Figure 2). for administering melatonin before bedtime is that it prevents
the soporific effects of melatonin from interfering with the
Enrolled participants subject’s usual activities during the day, and other trials have
During the study, 72 subjects were screened for participation, typically administered melatonin around bedtime, facilitating
33 were excluded, and 39 were randomized (Figure 1). The the comparison of study results.
demographic and metabolic baseline characteristics of the The doses of melatonin used in human clinical trials
39 enrolled subjects are shown (Table 2). The mean age at has ranged from 0.3 mg/day to 10.0 mg/day.9,13,14 The dura-
enrollment was 58.9 years. Over half of the subjects were tion of melatonin clinical trials has ranged from one-time
women, and 33% were African-American. administration, including blood pressure and sleep trials,
to daily dosing for up to 12 months.14 However, the typical
Treatment handling, dosage, duration, duration is 1–4 weeks. The two most significant concerns
rationale, toxicity, and safety regarding melatonin supplementation are: causing sleepiness
Both the study drug and placebo were supplied by the Inves- in some people with daytime doses of 8.0 mg/day, and giving
tigational Drug Service of Emory University. Melatonin melatonin at a time when it might cause an unwanted sleep/
powder was obtained from Professional Compounding wake phase shift. The list of side effects that we monitored
Centers of America, which obtains melatonin from Marcor included early morning wakening (a possible sign of a small
Development Corporation. Certif icates of purity were phase advance), headaches, dizziness, nausea, and drowsi-
provided for each batch of melatonin. Placebo capsules ness.14 In a meta-analysis addressing the efficacy and safety of
were compounded from lactose powder and were identical exogenous melatonin in 487 subjects, the distribution of these
in appearance to the melatonin capsules. All compound- potential side effects was similar between the melatonin and
ing was completed by the Investigational Drug Service placebo groups for doses up to 7.5 mg/day and for durations
of Emory University Hospital, which stored the drug and up to 12 months.14 Based on these prior studies, we deemed
placebo and dispensed prescriptions to subjects enrolled in an 8.0 mg daily dose to be safe and tolerable, and a 10-week

Placebo Melatonin

10 weeks 10 weeks

Day 1 End of second

Day 15 Day 25 6 weeks
Screening visit 1 period
Screening visit 2 First baseline Washout period
(check eligibility and evaluation visit
(placebo run-in) visit Day 26
obtain consent)
Randomization

Melatonin Placeb

10 weeks o 10

weeks

End of first Second

period baseline visit
evaluation visit

Figure 2 Schematic of study design.

Open Access Journal of Clinical Trials 2013:5 submit your manuscript | www.dovepress.com
55
Dovepress
Terry et al Dovepress

Table 1 Power to detect within-subject differences in components of the metabolic syndrome after treatment with melatonin
compared with placebo using a crossover design, assuming that 30 subjects completed the trial
Metabolic Projected mean within- Projected within- Power to detect projected Power to detect projected
syndrome subject difference subject standard mean within subject- mean within subject-
component (melatonin – placebo) deviation difference (melatonin – difference (melatonin –
placebo): 0.05 placebo): 0.01
Fasting glucose 5.5 6.7 0.86 0.66
(mg/dL)
24-hour systolic blood 5.5 7.4 0.79 0.55
pressure (mmHg)
HDL cholesterol 3.6 4.7 0.82 0.59
(mg/dL)
Triglycerides (mg/dL) 7.5 10.0 0.80 0.56
Waist 4.0 5.4 0.79 0.55
circumference (cm)
Abbreviation: HDL, high-density lipoprotein cholesterol.

period to be a sufficient duration of time for the components after collection. Prior to sample collection, participants were
of the metabolic syndrome to be influenced. advised to fast overnight and to not eat breakfast. GLUCm
reagent is used in conjunction with the Synchron LX system
Retention and compliance and Synchron LX Aqua Cal 1 and 2 (Beckman Coulter, Ful-
At the initial interview visit, all potential subjects received lerton, CA, USA) to measure quantitative glucose levels.
an orientation to the study. Spouses and housemates were
encouraged to attend this orientation because they could cor- Blood pressure
roborate whether or not the study subjects took their medica- We measure 24-hour ambulatory blood pressure using a
tions as prescribed. An information booklet described how Spacelabs 90217 ambulatory blood pressure monitor. This
and when to take the study pills and what to do if problems monitor has been validated for the measurement of systolic
were experienced (eg, lost pills or side effects). The project and diastolic blood pressure.15 The monitor is programmed to
manager also provided participants with a subject log so that record blood pressure every 60 minutes for a 24-hour period.
they could document any unusual symptoms while they were In addition, during General Clinical Research Center visits
taking the study medication. Participants were also provided at the beginning and end of each of the 10-week treatment
with the addresses (with postcards) and phone numbers of periods, blood pressure was measured using a sphygmoma-
the project investigators and project manager. The project nometer 5 minutes after the participant had been seated, twice
manager also educated each subject about how and where to in each of the right and left arms.
store the study pills and what time of day to take them.
Triglycerides
Study outcomes Triglycerides were measured by the Emory University Hos-
The primary outcomes were levels of the following factors pital Core Laboratory from fresh plasma within 6 hours of
after 10 weeks of melatonin or placebo supplementation: collection. Triglycerides glycerophosphate oxidase reagent
plasma glucose, office and 24-hour ambulatory blood pres- is used in conjunction with the Synchron LX System, the
sure values, triglyceride levels, HDL cholesterol levels, and Synchron systems multicalibrator, and triglycerides-blanked
waist circumference. These levels were measured on four assay parameters (Beckman Coulter) to determine quantitative
occasions during the study, ie, at the start and end of the first triglyceride levels. Triglycerides glycerophosphate oxidase
10-week period, and then (following the 6-week washout reagent is used to measure the triglycerides concentration
period) again at the start and end of the second 10-week by a timed endpoint method. Triglycerides in the sample are
period (Figure 2). hydrolyzed to glycerol and free fatty acids by the action of
lipase. A sequence of three coupled enzymatic steps using
Fasting glucose glycerol kinase, glycerophosphate oxidase, and horseradish
Glucose was measured by the Emory University Hospital peroxidase causes the oxidative coupling of 3,5-dichloro-2-
Core Laboratory from fresh plasma samples within 6 hours hydroxybenzenesulfonic acid with 4-aminoantipyrine to form

56 submit your manuscript | www.dovepress.com

Open Access Journal of Clinical Trials 2013:5
Dovepress
 Dovepress Melatonin supplementation and metabolic syndrome

Table 2 Baseline characteristics of enrolled subjects (n 39)

Characteristic Mean SD or Median Minimum Maximum
n (%)
Age (years) 60.0 10.1 59.4 35.0 78.1
Gender
Male 17 (43.6%) – – –
Female 22 (56.4%)
Race
Caucasian 24 (61.5%) – – –
African American 13 (33.3%)
Asian 1 (2.6%)
Hispanic/Latino 1 (2.6%)
Waist circumference (cm)
Males 112.5 8.9 111.3 99.1 133.1
Females 103.9 10.2 102.9 88.4 132.1
Triglycerides (mg/dL) 181.8 168.6 140.1 55.0 1043.7
HDL cholesterol (mg/dL)
Males 36.7 6.4 34.4 28.1 50.2
Females 46.8 10.2 44.4 32.1 74.0
Blood pressure (mmHg)*
Systolic 126.1 16.9 123.5 89.5 161.3
Diastolic 77.1 11.3 74.0 57.5 103.8
Fasting glucose (mg/dL) 102.6 13.8 102.8 74.0 137.9
Note: *Calculated as the average of four measurements (two per arm).
Abbreviation: HDL, high-density lipoprotein cholesterol.

a red quinoneimine dye. The Synchron LX system monitors Nutrition Examination Survey.16 Briefly, this procedure involves
the change in absorbance at 520 nm just prior to addition of locating the lateral border of the ilium of the standing study sub-
lipase and for a fixed time interval after lipase addition. This ject, drawing a horizontal line just above the uppermost lateral
change in absorbance is directly proportional to the concentra- border of the right ilium, placing a measuring tape around the
tion of triglycerides in the sample and is used by the system trunk in a horizontal plane at the level marked on the right side
to calculate and express the triglyceride concentration. of the ilium, and measuring at the end of a normal expiration.
Measurements are recorded to the nearest 0.1 cm.
HDL cholesterol
HDL cholesterol is measured by the Emory University Other study variables
Hospital Core Laboratory from fresh plasma samples within Interview data
6 hours of collection. HDLD reagent is used in conjunction The project manager interviewed subjects at the baseline visit
with the Synchron LX system and Synchron systems lipid to obtain data on demographics, health history, medication
calibrator (Beckman Coulter) to determine quantitative use, and dietary and behavioral factors. A questionnaire was
HDL levels. HDLD reagent is used to measure cholesterol administered at the end of each 10-week period to assess
concentrations by a timed-endpoint method. The Synchron potential side effects of exposure to melatonin or placebo.
LX system automatically proportions the appropriate HDL We asked specific questions regarding subject experiences
cholesterol sample and reagent volumes into a cuvette. The of nausea, dizziness, fatigue, daytime drowsiness, early
ratio used is one part sample to 93 parts reagent. The sys- morning waking, and any other sleep disturbance, as well
tem monitors the change in absorbance at 560 nm, which is as any other unusual symptoms the participant may have
directly proportional to the concentration of cholesterol in experienced during the trial.
the sample and is used by the system to calculate and express
the HDL cholesterol concentration. Endogenous melatonin
Endogenous melatonin levels in overnight urine (first morning
Waist circumference void) samples will be measured using a radioimmunoassay
We measured waist circumference using the procedure (Alpco Diagnostic, Windham, NH, USA) with a sensitivity
employed by the 2003–2004 wave of the National Health and of 0.2 pg/mL. Differences in morning melatonin levels are

Open Access Journal of Clinical Trials 2013:5 submit your manuscript | www.dovepress.com
57
Dovepress
Terry et al Dovepress

representative of relative differences in night-time and overall test of whether melatonin improves at least one of the five
melatonin levels.17 Analyses will be performed to examine metabolic syndrome endpoints. The power calculations were
whether the effects of melatonin supplementation are modi- based on the multivariate Hotelling T2 module in PASS 2005
fied by endogenous melatonin level. (Kaysville, UT, USA). Based on a sample size of 30 subjects,
we achieved statistical power exceeding 90% for the global
Sleep duration and quality test, with a two-sided alpha of 0.05.
We assessed sleep duration and quality using two approaches:
a standard sleep log used by the Emory Program in Sleep, Data management
Aging, and Chronobiology18 and a conventional question- Microsoft Access, an American National Standards Institute
naire to measure sleep quality, ie, the Pittsburgh Sleep compliant relational database management system, was used
Quality Index.19 The sleep log collected daily data on total for data entry and management. The data entry screens mimic
sleep time. The Pittsburgh Sleep Quality Index, which has paper case report forms for efficient data entry. Double data
a one-month time frame, was also to be administered at the entry was utilized.
screening visit and three subsequent study visits, including
the final visit. Analyses were performed to examine whether Data analysis plan
the effects of melatonin supplementation may be acting in Analysis of primary endpoints was performed according to
part through changes in sleep duration and/or quality. the intent-to-treat principle. Standard statistical methods for a
one-period crossover design were employed using Statistical
Oxidative stress and inflammation biomarkers Analysis Software (SAS Inc, Cary, NC, USA). This includes
To assess potential underlying biologic mechanisms, we are testing for both a carryover effect and an order effect. To assess
assessing blood levels of several key markers of inflammation the possibility that subjects were aware of their treatment order,
and oxidative stress, including high-sensitive C-reactive protein, we collected data at the end of treatment on whether each sub-
tumor necrosis factor-alpha, interleukin-6, F2-isoprostanes, and ject thought he or she was first randomized to take melatonin or
oxidized low-density lipoprotein cholesterol. placebo. Dropouts were small in number and not deemed to be
related to the study treatment. Specifically, two subjects dropped
Sample size and power calculations out during follow-up: one male participant dropped out upon his
We conducted statistical power calculations to assess the physician’s advice in order to receive calcium channel block-
ability of our trial to detect moderate, clinically meaning- ing medication, and another male participant self-withdrew
ful improvements in our primary endpoints in this Phase II (without explanation) during the second of the two 10-week
crossover trial. We planned to recruit 40 subjects with at least treatments, when the participant was taking placebo. These two
30 of them completing the trial. Therefore, we calculated subjects with partial data are included in the intent-to-treat data
power for a sample of 30 subjects (Table 1). To estimate analysis of the primary and secondary endpoints.
standard deviations, we used values of the between-subject
standard deviations for each of our primary endpoints as Discussion
reported in previous studies of metabolic syndrome.20–22 Melatonin supplements are widely used in the US, with many
We assumed conservatively that the standard deviations in millions of dollars in annual sales. Understanding the effects
our study would be 66% of the between-subject standard of this widely used supplement on metabolic conditions that
deviations from these previous studies. For triglycerides, we affect many individuals is an important public health issue.
based our estimates on the likely variability on the natural Although most of the literature suggests that melatonin is
log-transformed values because these data are typically right- safe and not associated with any major side effects,14 few studies
skewed. After making these assumptions, we found that we have examined the effect of melatonin supplementation for the
would have approximately 80% statistical power to detect duration that we have planned. Although studies that have con-
each moderate effect size using a two-sided type I error ducted long-term trials and/or with higher doses of melatonin
rate (alpha) of 5% based on 30 subjects who completed the have not reported serious side effects, subjects in our study were
crossover design (Table 1). If we use a Bonferroni adjustment monitored carefully for any potential adverse effects.
alpha of 0.05/5 0.01, then the statistical power is reduced to We chose a crossover design for the MetSyn trial rather
approximately 55%–65% for each of the five comparisons. than a parallel group design in order to maximize effi-
We also calculated the statistical power for the “global” ciency.11 We assumed that neither a carryover effect nor an

58 submit your manuscript | www.dovepress.com

Open Access Journal of Clinical Trials 2013:5
Dovepress
 Dovepress Melatonin supplementation and metabolic syndrome

order-of-treatment effect was likely, nor was a significant 2. Davis AA, Kaklamani VG. Metabolic syndrome and triple-negative
breast cancer: a new paradigm. Int J Breast Cancer. 2012;2012:
loss to follow-up over the 26-week period of the trial. As 809291.
described above, a placebo run-in period of ten days prior to 3. Doyle SL, Donohoe CL, Lysaght J, Reynolds JV. Visceral obesity,
randomization (to assess compliance) and a 6-week washout metabolic syndrome, insulin resistance and cancer. Proc Nutr Soc.
2012;71:181–189.
period between the first treatment period and the second 4. Pevet P, Challet E. Melatonin: both master clock output and internal
treatment period (to eliminate a possible carryover effect) time-giver in the circadian clocks network. J Physiol Paris. 2011;105:
170–182.
were key design features of this trial. 5. Grossman E, Laudon M, Zisapel N. Effect of melatonin on nocturnal
On the whole, evidence from animal experiments, blood pressure: meta-analysis of randomized controlled trials. Vasc
observational epidemiologic studies, and a previous clinical Health Risk Manag. 2011;7:577–584.
6. Peschke E, Bach AG, Mühlbauer E. Parallel signaling pathways of mela-
trial suggest that melatonin supplementation may protect tonin in the pancreatic beta-cell. J Pineal Res. 2006;40:184–191.
against elevated glucose/insulin resistance, elevated blood 7. Ríos-Lugo MJ, Cano P, Jiménez-Ortega V, et al. Melatonin effect
on plasma adiponectin, leptin, insulin, glucose, triglycerides
pressure, dyslipidemia, and perhaps abdominal obesity.5–9 and cholesterol in normal and high fat-fed rats. J Pineal Res.
However, the evidence is limited in part due to the lack 2010;49:342–348.
of randomized clinical trials in humans.9 There have been 8. Nduhirabandi F, du Toit EF, Lochner A. Melatonin and the meta-
bolic syndrome: a tool for effective therapy in obesity-associated
few clinical trials of melatonin supplementation to assess abnormalities? Acta Physiol (Oxf). 2012;205:209–223.
effects on glucose levels, blood lipid levels, or body weight 9. Koziróg M, Poliwczak AR, Duchnowicz P, Koter-Michalak M, Sikora J,
Broncel M. Melatonin treatment improves blood pressure, lipid profile,
and/or size. Although there have been more clinical trials of and parameters of oxidative stress in patients with metabolic syndrome.
melatonin supplementation focused on blood pressure, most J Pineal Res. 2011;50:261–266.
10. Schernhammer ES, Kroenke CH, Dowsett M, Folkerd E, Hankin-
have been of small sample size and their results have been
son SE. Urinary 6-sulfatoxymelatonin levels and their correlations
inconsistent. This trial will provide knowledge of the effects with lifestyle factors and steroid hormone levels. J Pineal Res.
of melatonin in subjects with metabolic syndrome, and lay the 2006;40:116–124.
11. Louis TA, Lavori PW, Bailar JC 3rd, Polansky M. Crossover and
groundwork for future clinical trials of melatonin in subjects self-controlled designs in clinical research. N Engl J Med. 1984;310:
with metabolic syndrome. 24–31.
12. Lusardi P, Piazza E, Fogari R. Cardiovascular effects of melatonin in
Acknowledgment hypertensive patients well controlled by nifedipine: a 24-hour study.
Br J Clin Pharmacol. 2000;49:423–427.
This trial is approved by the Emory University institu- 13. Brzezinski A, Vangel MG, Wur tman RJ, et al. Effects of
tional review board (14784), has Investigational New exogenous melatonin on sleep: a meta-analysis. Sleep Med Rev. 2005;
9:41–50.
Drug approval by the US Food and Drug Administration 14. Buscemi N, Vandermeer B, Hooton N, et al. Efficacy and safety of
(105764), and is registered at ClinicalTrials.Gov exogenous melatonin for secondary sleep disorders and sleep disor-
ders accompanying sleep restriction: meta-analysis. BMJ. 2006;332:
(01038921). The data safety monitoring board comprised 385–393.
William McClellan, Sol Jacobs, and Janice Lea, Emory 15. Baumgart P, Kamp J. Accuracy of the SpaceLabs Medical 90217
University School of Medicine. The study was funded by the ambulatory blood pressure monitor. Blood Press Monit. 1998;3:
303–307.
National Institutes of Health/National Center for Comple- 16. Centers for Disease Control. NHANES 2003–2004 Documentation.
mentary and Alternative Medicine R21 AT004220-01A2, Available from: http://www.cdc.gov/nchs/data/nhanes/nhanes_03_04/
BM.pdf. Accessed 29 January, 2013.
and was supported by the National Center for Advancing 17. Graham C, Cook MR, Kavet R, Sastre A, Smith DK. Prediction of
Translational Sciences of the National Institutes of Health nocturnal plasma melatonin from morning urinary measures. J Pineal
under Award Number UL1TR000454. The content is solely Res. 1998;24:230–238.
18. Bliwise DL, Ansari FP, Straight LB, Parker KP. Age changes in timing
the responsibility of the authors and does not necessarily and 24-hour distribution of self-reported sleep. Am J Geriatr Psychiatry.
represent the official views of the National Institutes of 2005;13:1077–1082.
19. Buysse DJ, Reynolds CF 3rd, Monk TH, Berman SR, Kupfer DJ. The
Health. Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice
and research. Psychiatry Res. 1989;28:193–213.
Disclosure 20. Cuspidi C, Meani S, Fusi V, et al. Metabolic syndrome and target organ
damage in untreated essential hypertensives. J Hypertens. 2004;22:
The authors report no conflicts of interest in this work.
1991–1998.
21. Desai MY, Dalal D, Santos RD, Carvalho JA, Nasir K, Blumenthal RS.
References Association of body mass index, metabolic syndrome, and leukocyte
1. Grundy SM, Brewer HB Jr, Cleeman JI, Smith SC Jr, Lenfant C; count. Am J Cardiol. 2006;97:835–838.
American Heart Association, National Heart, Lung, and Blood Institute. 22. Duggirala R, Blangero J, Almasy L, et al. A major locus for fasting
Definition of metabolic syndrome: Report of the National Heart, Lung, insulin concentrations and insulin resistance on chromosome 6q with
and Blood Institute/American Heart Association conference on scientific strong pleiotropic effects on obesity-related phenotypes in nondiabetic
issues related to definition. Circulation. 2004;109:433–438. Mexican Americans. Am J Hum Genet. 2001;68:1149–1164.

Open Access Journal of Clinical Trials 2013:5 submit your manuscript | www.dovepress.com
59
Dovepress
Terry et al Dovepress

Open Access Journal of Clinical Trials Dovepress

Publish your work in this journal
The Open Access Journal of Clinical Trials is an international, peer- methodologies. The manuscript management system is completely
reviewed, open access journal publishing original research, reports, online and includes a very quick and fair peer-review system, which
editorials, reviews and commentaries on all aspects of clinical trial is all easy to use. Visit http://www.dovepress.com/testimonials.php to
design, management, legal, ethical and regulatory issues, case record read real quotes from published authors.
form design, data collection, quality assurance and data auditing
Submit your manuscript here: http://www.dovepress.com/open-access-journal-of-clinical-trials-journal

60 submit your manuscript | www.dovepress.com

Open Access Journal of Clinical Trials 2013:5
Dovepress