~ Freund Publishing House Ltd., London

----.-

Journal of Pediatric Endocrinology & Metabolism. 12. 57-67 (1999)

Effects of a Low Dose of Melatonin on Sleep in Children withAngelman Syndrome

Irina V. Zhdanova, Richard 1. Wurtman and Joseph Wagstaft'*

Department of Brain and Cognitive Sciences, MIT, Cambridge, MA.Genetics Division, Children's Hospital, Boston, MA, USA

ABSTRACt

The effects of low dose melatonin therapy onsleep behavior and serum melatonin levels werestudied in Angelman syndrome (AS) childrensuffering from insomnia. 24-hour motor activitywas monitored in 13 AS children (age 2-10 yr) intheir home environments for 7 days prior tomelatonin treatment and for 5 days duringwhich a 0.3 mg dose of melatonin was admin-istered daily 0.5-1 hour before the patient'shabitual bedtime. Blood samples were with-drawn at hourly intervals over two 21-hourperiods in order to measure individual endogen-ous serum melatonin levels and the levelsinduced by melatonin treatment. Actigraphicrecording of motor activity, confirmed byparents' reports, showed a significant improve-ment in the patients' nocturnal sleep pattern asa result of melatonin treatment. Analysis of thegroup data revealed a significant decrease inmotor activity during the total sleep periodfollowing melatonin treatment, and an increasein the duration of the total sleep period. Endo-genous peak nocturnal melatonin values rangedfrom 19 to 177 pg/mL The administration ofmelatonin elevated peak serum hormone levelsto 128-2800 pg/ml in children of different agesand body mass. These data suggest that amoderate increase in circulating melatonin levelssignificandy reduces motor activity during thesleep period in Angelman syndrome children,and promotes sleep.

Reprint address:Irina V. Zhdanova, M.D., Ph.D.E18-439MIT50 Ames St.Cambridge. MA 02139, USA

VOLUME12.NO.1, 1999

------

INTRODUCl'ION

Angelman syndrome (AS) is a rare geneticdisorder (incidenceestimatedto be approximately1in 20,000) characterized by severe mental retard-ation with absent speech, seizures, ataxia, charact-

eristic facial features with easily provoked smilingand laughter, and disturbed sleep. About 70% ofAS individuals have a de novo deletion ofchromosome 15 bands q11-13 on the chromosomeinherited trom their mothe~. Three other etiologictypes of AS include paternal uniparentaldisomy 15(2-5%), imprintingmutations (2-5%) and AS casesthat have no evidence of deletion, uniparentaldisomy, or imprinting mutations (20-25%). In thislast group, mutations in the UBE3A1E6-APgenehave recently been described3.4.

The most common behavioral problems in ASchildren are hyperactivity, attention deficit anddifficulties initiating and maintaining sleep. Sleepproblemsare usuallydetectable at severalmonthsofage and persist for many years. Long latencies tosleep and prolonged awakenings at night lead tofatigue and sleepinessupon morningawakeningandcause a chronic sleep debt which may potentiateother behavioraland neurologicproblems.

Observations in work with human babies reveala correlation between the timing of the consolid-ation of nocturnal sleep and the normal onset ofrhythmic melatonin secretion, both of which occurwhen the infant is about 3 months 0Ids.6.Likewise,the concurrent decline in melatonin secretion andsleep efficiencywith age are thoughi to be relatedphenomena; for example, middle-aged and elderlyinsomniacs reportedly exhibit lower melatoninproduction than do good sleepers of the sameage7.8.Initial studies regarding the' acute effects ofmelatonin in humans revealed that pharmacologicaldoses of the hormone induce sleepinessand sleep9-11.Recently it has been shown that low melatonin

57

58LV. ZHDANOVA ET AL.

'.

doses (0.1-0.3 mg), which induce serum hormoneconcentrations comparable to those typical ofnoctumalmelatoninlevels in adults, are sufficienttofacilitate sleep induction in healthy young maleswhen the hormone is administeredat differenttimesof the day, from noon to 9 pmI2-14.Clinicalstudiesrevealed that administration of pharmacologicaldoses of melatonin (2.5-7.5 mg, p.o.) to multiplydisabled children with severe sleep disorders, whohad failed to respond to conventional manage-ment1', or to children suffering from Rett syn-drome'6, substantiallyimproved their sleep patternsand increased sleep duration according to care-givers' reports, or decreased their latency to sleeponset as assessed actigraphically.

We tested the possibility that actigraphicallyregisteredmotor activityduring sleep in AS childrenwould be diminished when their circulating mela-tonin levels were increased to within the physiol-ogical or low pharmacological range. This wasaccomplished by the administration of a 0.3 mgdose of melatonin to the patients close to theirbedtime.

PATIENTSANDMEmODS

Patients were recruited from the population ofAS children diagnosed or treated at Boston Child-ren's Hospital, or whose parents responded toinformation about the study distributed by theAngelmanSyndrome Foundation. The investigationprotocol was approved by the Children's Hospitalhuman research committee. Parents signed aninformed consent form prior to their child'sparticipationin the study. Thirteen children, aged 2-10 years old at the time of entry into the study,participated in the protocol. Twelve patients hadtypical 15qll-q 13 deletions, confirmed by eitherFISH analysisor Southern blot analysis,and one (#8) had paternal UPD 15 resulting from a 13;15Robertsonian translocation inherited along with anormal chromosome 15 from the father. All patientsshowed typical clinical features of AS includingsevere developmentaldelay with absent or minimalspeech, and seizures. Patient # 3 had been treatedwith melatonin(purchased in a health food store) 3mg daily for approximately 4 months; this wasdiscontinued 2 weeks prior to entering the study.

Patient # 8 had been treated with 3-9 mg ofmelatonin each night for 4 months, and this too wasdiscontinued 2 weeks prior to entry into the study.Patient # 4 had been treated with chloral hydrate,500 mg each night, for approximately 3 years; thistreatment also was discontinued 2 weeks before

entering the study. Patients #10 and 11 are mono-zygotic twins. The children's gender, weight, age,and initial medications are summarized in Table I.

The patients' sleep/wake schedules were monit-ored in their home environments for 7 days prior tomelatonin treatment (baseline) and for 5 days duringwhich a 0.3 mg dose of melatonin was administeredY2-1 hour before the patient's habitualbedtime. Thepatient's parents maintained a sleep diary for theirchild, indicating bedtime, approximate times offalling asleep, and times of awakenings. Objectivedata on 24 hour motor activities were obtainedusing a portable body actigraph (Mini-logger2000,Mini Mitter, Oregon), worn by the patient in apocket on the back of a cloth vest. The recordermonitored the number of body movements perminute. On the first day and night after the bloodwithdrawal session the actigraphic recording wasnot obtainedbecause the child's sleep was disturbedon the night the blood was drawn. That circum-stance could provoke a rebound increase in sleepduring the following night, thus distorting therecord of the treatment period. The sleep diarymaintained by patient's parents described periodswhen the actigraphwas not worn.

After seven days of a baselinemotor activityandsleep/wake assessment, patients were admitted tothe Children's Hospital ClinicalResearch Center toevaluate their endogenous melatonin secretion patt-erns. Patients' rooms were maintainedat 72°F, andlights were dimmed at 19.00 h to <30 lux. Bloodsamples (2 ml each) were drawn hourly from 15.00h to 10.00 h the next morning, through an intra-venous catheter inserted into a forearm vein; aheparin lock was used to prevent clotting. Serumsamples,separated by centrifugation,were stored at-20°C until assayed for melatonin concentration.The patient's core body temperatures were monit-ored using a rectal probe (YSI, Yellow Springs,Ohio) and recorded by a Mini-logger2000. Patientsreceived regular meals through the day which weresimilarto their habitualdiet.

JOURNALOFPEDlAlRICENDOCRINOLOGY&:METABOLISM

Starting on the second day after the initialhospital admission,patients received a 0.3 mg oratdose of melatonin on each of six consecutiveevenings,Yz-l h prior to their habitualbedtimeranging trom 19.30 to 20.30 h. The contents of agelatin capsule (0.3 mg melatonin, Nestle Co.,Switzerland,and microcrystallinecellulose 'Avicel')were mixed with a teaspoon of a semi-liquidfood(e.g., pudding or applesauce) and added to thechild's evening snack. Each patient's activity andsleep were assessed as described above. After sixdays of melatonin treatment, patients were againadmitted to the Clinical Research Center (CRe),and blood samples were drawn hourly trom 17.00

VOLUME12,NO. I, 1999

to 11.00 h the followingmorning in order to assessthe effects of exogenous melatonin (0.3 mg doseadministered at 19.30 h) on their serum hormoneprofiles.The environmentand procedures were heldsimilar to .those used during the first CRCadmission.

Patients' motor activities,recorded by actigraph,were analyzedusing a software algorithmdevelopedat MIT. The total sleep period (TSP) within abedtime period was defined as an interval elapsingbetween the first of ten consecutiveminuteswithoutmovements, and the last minute of the last tenminute interval without movements.We chose notto use standard actigraphic algorithm criteria

MELATONIN AND SLEEP IN ANGELMAN SYNDROME 59

TABLEt

Patients' demographics and medication during the study

Age Gender Weight Medications

9 F 29.5 kg Tranxene (375 mg qAM, 562 mg qhs)

Vitamine E (400 IU qhs)

2 7 M 30.9 kg Depakene (300 mg q AM,250 mg q3PM, 250 mg qhs)

3 6 F 21.5 kg Depakote (250 mg qAM, 125 mg q3PM, 250 mg qhs)

4 10 M 21.4 kg Depakote (375 mg qAM, 375 mg q12N, 375 mg qhs)

Desipramine (10 mg qAM, 30 mg qhs)

Mebaral (50 mg qAM, 50 mg qhs)

5 7 M 23.9 kg Neurontin (100 mg qAM, 200 mg qhs)

Zarontin(125mgqAM,125mgqhs)

6 7 F 24.6 kg Klonopin (1 mg qAM, 1 mg q3PM, 1.5 mg qhs)

7 7 F 27.9 kg Valium (2.5 mg qAM)

8 10 F 60.2 kg Depakote (500 mg bid), Clonidine(0.1mgbid)

9 4 F 17.9 kg Mogadon(2.5 mg qhs), Depakote(500 mgbid)

10 2 F 9.8 kg Valproi acid (245 mg lid)

11 2 F 10.1kg Valproicacid (245 mg lid)

12 4 F 15.8 kg Depakote(250 mg qAM, 125mg q3PM,250 mg qhs)

13 10 M 30.9 kg Depakote(312 mg tid), Phenobarbital(45 mg qhs)

._.u_ .__

60 LV. ZHDANOVA ET AL.

differentiating wakefulness and sleep within thesleep period since we lacked parallel polysomno-graphic data in work with this population. Becausewe did not have objective data on the timing oflights out in the home environment,we question theprecision of our estimates of latency to sleep onsetand did not include this parameter in our statisticalanalysis of the experimental data obtained. Thus,analysisof actigraphicdata was limitedto TSP andto the number of movements per hour during theTSP (M).

Melatonin concentrations were measured in 0.5ml serum aliquots using a commercially availableradioimmunoassay kit (Buhlmann Laboratories;Allschwil, Switzerland). Extraction was accomp-lished using e 18 columns.The limitof detection ofthe assay was 2.2 pmol/1(0.5 pglml). The intra-assay coefficients of variation for control sampleswere 7.2% (38.8 pmol/1[9 pglml]) and 7.8% (94.8pmol/1 [22 pglml]); the corresponding inter-assaycoefficients of variation were 12.6% and 16.1%.The parameters of interest were time of onset andoffset of nocturnal melatonin secretion, area underthe time-melatoninconcentration curve within thisperiod (AUC), and peak night-time and minimumdaytime hormone levels. The onset and offset ofmelatonin production were defined as the timepoints at which evening or morning serummelatonin reached a concentration two standarddeviationsabove the mean daytime level. AUe wasmeasured between the "onset" and "offset" timepoints. If, for technical reasons, blood withdrawalwas interrupted close, but prior to an estimated"offset" time, as indicated in Table 2, AUe wasmeasured within the period for which data wereavailable.

Parents were given the option, as an extensionofthis study, to continue melatonintreatment of theirAS childrenfor up to one year, using either a 0.3 ora 0.2 mg dose of melatoninas a continuationof thedescribed study. Twelve of the thirteen familieschose to continue the treatment. In order toevaluate a possible shift in the phase of melatoninsecretion, we repeated an overnight blood with-drawal in three children (# I, 2 and 5) after several

. weeks of melatonintreatment.Statistical analyses of the actigraph and mela-

tonin concentration data involved repeated meas-ures ANOVA to evaluate the differences between

baseline levels of outcome variables and the levelon a 0.3 mg melatonintreatment. Due to schedulinSor technical problems, not all of the patients' da~sets contained an equal number of baseline and/ortreatment day recordings. Individual and groupmeans represent four consecutive days of a baselineperiod and four consecutive days of a treatmentperiod. In the cases where data were missing Wechose to analyze any available four consecutivedays, otherwise the last four days of the period wereused for the analysis.

In four of the children (# 6,9, 10, II) who hadto move trom their homes to Boston for bloodwithdrawal sessions, we chose to analyze activityduring four consecutive days of melatonin treatmentwhile they were back at home, rather than therecordings obtained during their stay in a hotel.

RESULTS

Endogenous serum melatonin profiles for thegroup of AS children studied were highly variablewith respect to peak melatonin levels and areasunder the time-concentration curves (AUe, Table2). Peak nocturnal values ranged from 81.9 pmoVl(19 pglml) to 762.9 pmoVI(177 pg/ml), the groupmean value (:i:SEM) was 387.9:i:63.4 pmoVI(90%14.72pglml) (Fig. I). Melatonin AUe valuesranged trom 607.7 pmvtll'h (l41 pg/ml'h) to 4775pmol/1'h(1108 pglml'h) in different patients, groupmean value (:i:SEM) was 2745:i:429.3 pmoVl'h(637:i:99.61pglml'h). In ten of the children the on-set of nocturnal hormone secretion was consistentwith their bedtime and their sleep onsets; howeverin three of the patients (# I, 2 and 13) the nocturnalsurge was substantiallydelayed (Table 2). Prior totreatment, the sleep pattern in these three childrendid not correlate with the timing of melatoninsecretion, i.e., their habitual bedtime and their sleeponset usually occurred in advance of the onset innocturnal melatonin release. However, in two ofthem (# I and 13), daytime sleepiness untilapproximatelynoon was described by their parents,prior to melatoninadministration.

Symptoms of sleep disturbance, as reported byparents, were variable among the patients, includinglong periods of activity after the lights were out,

JOURNAL OF PEDIATRIC ENDOCRINOLOGY & METABOLISM

several prolonged awakenings at night with highmotor activity, and early morning awakenings.Lowsleep quantity and quality, reported by parents, wasconfirmed by actigraph data, collected prior to thetreatment, showing a high number of movementsduring a sleep period (Table 3). In our age-diversegroup of children we did not find. a significantcorrelation between an individual's endogenouspeak melatonin levels or AUe and the number ofhis or her movements per hour of TSP underbaselineconditions.

Within an hour after ingestion of a capsulecontaining 0.3 mg of melatonin, circulating melat-onin levels increased significantly, and remainedabove the daytime baseline levels for 12-15 hours(Fig. 1). The administration of a uniform 0.3 mg

VOLUME 12, NO. I, 1999

dose of melatonin elevated peak serum hormonelevels to 552-12068 pmol/l (128-2800 pg/ml) inchildrenof different ages and body mass (Table 2).Mean peak circulating melatonin levels after thetreatment were 2701.2:1:1038pmol/l (601.6:%:223pg/ml) (Fig. 1), significantlyhigher (p<O.OOI)thanthe mean peak endogenous melatoninlevel.Highestpeak levels following treatment were in the twoyear-old identical twin patients (# 10 and 11),whosebody masseswere the lowest in the groupstudied (Table I). The mean group AUe value wassignificantly increased (p<O.OOI) following theadministration of a 0.3 mg dose of melatonin(11564:%:3276pmol/l'h [2683:%:759.98pg/ml'h];Table 2). Daytime melatonin levels were Ie ; than21.6 pmol/l (5 pg/ml) for all the children studied

MELATONIN AND SLEEP IN ANGELMAN SYNDROME 61

TABLE 2

Serum melatonin patterns in AS children prior to treatment (0 mg)and as a result of melatonin treatment (0.3 mg)

Melatonin secretion

Peak levelpglml "Onser' "Offser' "Offser' AUC AUC

Omg 0.3 mg Omg o mg 0.3 mg Omg 0.3 mg

177 300 24.00h .- 11.00 h 888 "'2611

2 19 165 24.00 h 11:00 h 12.00 h 141 1070

3 49 320 22.00 h 08.00 h . 350

4 50 96 21.00 h 11.00 h 11.00 h 430 629

5 65 179 22.00 h 08.00 h 08.00 h 453 1117

6 129 197 22.00 h 09.00 h 10.00 h 885 1334

7 120 420 21.00 h 11.00 h 12.00 h 1023 2279

8 132 690 22.00 h 09.00 h - 1008 "3966

9 54 276 21.00 h 10:00 h 11.00 h 492 1538

10 148 2800 20.00 h 09.00 h 11.00 h 1040 9807

11 148 1850 20.00 h 09.00 h 12.00 h 1108 5312

12 41 400 21.00 h 10:00 h 10.00 h 238 1693

13 38 128 01.00 h 1100 h 10.00 h 228 845

'Samples not available after: '03.00 h; "07:30 h;'" 11.00 h;

and were not significantlychangedafter fivedays ofmelatonintreatment.

Administration of a 0.3 mg dose of melatoninsubstantially reduced the measured night-timemotor activityin eleven of the patients studied. Thisfindingwas consistent with parents' reports regard-ing the children's sleep quality in response to treat-ment (Table 3). However, nocturnal motor activitydid not change in two of the patients (# 4 and 12)whose baseline activities did not exceed 30 move-ments per hour. Analysisof the group data revealeda significantdecrease (p<O.001) in motor activity.M, duringthe total sleep period followingmelatonintreatment (Fig. 2), and a significantincrease in theTSP (p<0.05). The magnitude of the observedeffect on nocturnal motor activity did not signifi-cantly depend on the peak melatoninlevelsafter the

0.0 0.3

Fig,1:Melatonin Dose (mg)

Mean (:SEM) peak serom melatonin levels inthirteenAS children:endogenouslevels(0.0) andaftertheadministrationof a 0.3 mgdose.

--

JOURNAL OF PEDIATRIC ENDOCRINOLOGY&. METABOLlStl.l

62LV. ZHDANOVAETAL

TABLEJ

Actigraphically recorded total sleep period (TSP)and the number of movements per hour ofTSP (M)

no treatment 0.3 mg melatonin

TSP SEM M SEM TSP SEM M SEM

587.2 79.9 143.9 45.7 614 49.3 9.3 18

2 524.5 104.6 60.4 20.7 697 53.3 15.9 6.7

3 558.5 21.2 66.2 20.8 575.7 30.3 9.4 2.1

4 539.7 19.8 28.2 19.2 572.2 13.6 20.4 7.4

5 450 25.5 64.2 19.1 643.7 32.1 2.5 1.53

6 485.7 27.2 374.3 130.9 593.7 18.5 11.4 3.5

7 617.2 39.8 72.7 21.5 550.5 23.1 9.4 1.9

8 524 54.2 301.2 86.2 680.7 17.9 58.6 14.1

9 471 32.2 89.1 35.9 528.7 44.1 50.3 245

10 706.2 35.5 237.6 78.9 608.2 33.2 25.6 6.9

11 644 29.9 81.6 38.6 580.2 34.6 12.6 2.6

12 475 31.3 17.9 11 540 18.7 10.5 6.3

13 628.7 40.61 113.6 15.75 587.5 11.55 8.2 1.57-

.Missing baseline data was substituted using data collected after a 3 week melatonin withdrawalSEM -standard error of the mean

-800

CICI.-c'2 6000-IIIa;E 400E...G) 200II)

IIIG)Q.

0

. - .. .. .. -- .'. .....-.

MELATONIN AND SLEEP IN ANGEL MAN SYNDROME

150

..:=o.cUi 100-cCDECD>oE 50>-~om

o0.0 0.3

Melatonin Dose (mg)

Fig. 2: Mean (:t:SEM)motor activity per hour of the sleepperiod in thirteen AS children before and duringmelatonin tteatment

treatment, nor on the difference between theendogenous level of the hormone and serummelatoninconcentrations inducedby the treatment.

During the one week period of melatoninadmin-istration at home, parents reported that patientsusually fell asleep within 15 to 60 minutes aftermelatonin administration, and that sleep onsetoccurred faster than it did prior to the treatment. Insome cases, parents felt that the children, afterfalling asleep, were not aroused by sounds thatwould have awakened them prior to starting treat-ment. Parents reported that children were alertduring the day after melatonin administration, andthat there was either no change in daytimebehavioror that the patients were more attentive andinteractive.

Core body temperature recordings were notaccurate due to frequent displacement of the rectaltemperature probe, and these measurement errorswere highly provoked by the blood withdrawalprocedures. Thus, analysis of these temperaturerecordings is not presented.

In patients # 1, 2 and 13 the pretreatment onsetof nocturnal melatonin secretion was substantiallydelayed (Table 2). Administrationof a low dose ofthe hormone an hour prior to bedtime for twoweeks (patient # 1, Fig. 3C) or four weeks (patient

VOLUME 12, NO.1, 1999

--- - -----

63

# 2) ~roduc~d a 2 or 3 hour phase advance,r~spect~vely, 1Dthe onset of melatonin secretion.FIg. 3 Illustrates the sleep patterns in patient # Iduring melatonin treatment (A) and after treatmentwas stopped (B). Withdrawal from treatmentresulted in a delay in the sleep onset, which startedto occur close to the time of the 'new', shifted,onset of her melatonin secretion around 23.00 h. Itis interesting to note that her typical pretreatmentsleep pattern, according to her mother's report, wascharacterized by a 2-3 hour sleep period startinglate in the evening, and a 2-3 hour sleep period earlyin the morning. Thus, while the patient's initial sleepschedule was not harmonized with her nocturnalmelatonin secretion, melatonin treatment synchron-ized the two patterns. In contrast, administration of

a 0.3 mg dose for 4 weeks followed by a 0.2 mgdose for 4 weeks to patient # 5, whose initialmelatonin secretion onset was in phase with hishabitual bedtime, caused no temporal or quantit-ative change in his nocturnal melatonin production(Fig. 4C), in spite of a significant sleep promotingeffect of the treatment (Fig. 4B).

DISCUSSION

The results of this study suggest that an inducedmoderate increase in circulating melatonin levels inchildren afflicted with Angelman syndrome pro-motes regularized and less interrupted sleep.Continuous actigraphic recording documented asignificant decrease in motor activity during thesleep period while the children were treated with adaily low dose of melatonin. In addition, accordingto their parents' reports, most of the childrenshowed signs of fatigue within 15 to 30 min afterthe treatment was administered, their latencies tosleep onset were shortened, and their sleep wasmore consolidated.

Neurologically disabled and mentally retardedchildren frequently exhibit disrupted sleep/wakecycles and severe sleep disturbancesl7,18which aredifficult to treat using traditional sedatives. Insom-nia in such a population may be related to ab-normalities in brain development, which mayinclude abnormal development or function of thecircadian system, a decrease in the expression ofmelatonin receptors, a reduction in melatonin

Fig.J:

o17 18 19 20 21 22 23 24

HOURSSleep onset, judged from motor activity, in patient # 1 (A) during melatonin treatment, and (B) during melatoninwithdrawal after 4 weeks of treatment; (C) timing of the onset of elevated serom melatonin ~ priorto treatment,(.)during treatment, and (8) after 4 weeks of melatonin treatment.

-eCIS:c'2o-IV'iee:J..CDen

1

11

production, or a change in the brain's sensitivitytothis pinealhormone.

It has been shown that newborn infants do notdisplay rhythmic melatonin secretion for the firsttwo to three months6 but rely on the hormonesupplied via the mother's milk. Total melatoninproduction increases rapidly during the first year of

life, and might be important for brain developmentand maturation of the circadian system, includingsleep/wake mechanisms. The highest night-timemelatoninlevels have been observed in very youngchildren, aged 1-3 years (329.5%42 pglml)19.Melatonin levels start to fall around the time of theonset of puberty. A recent study2°also reported a

JOURNAL OF PEDIATRIC ENDOCRINOLOGY &; METABOLISM

--- _.~~ -~ --- --- ----.

64 LV ZHDANOVA ET AL.

200 .A

I i I150

100

50='5.......<II-=

e200

I I BI Ire>. 150

"C

=100

50

24 bour cycles

200 l- t C150 l- I

I100 I- ,I50 l- I

MELATONIN AND SLEEP IN ANGELMAN SYNDROME 65

150A

100

16 18 20 22 24 2 4 6 8 10

HOURS

Fig.4: Motoractivityof patient# S during(A)fourdaysofno treatment,and (B)duringfourdaysof melatonintreatment;(C)profilesof serummelatoninconcentrationspriorto treatment(A),duringtreatment(C), and after4 weeksof melatonintreatment(0).

substantial interindividual variability in circulatingmelatoninlevels,which decline from 175%109 pglmlto 128%44pglml during the course of puberty ingroups of childrenfrom 5 to 17 years of age. It wasnot possiblefor us to recruit a group of AS childrenof a homogeneous age, nor a matched controlgroup of children to study. Normative values for

circulating melatonin concentrations have not beenestablished for any age group. The majority of ourpatients, however, exhibited peak melatonin levelslower than the published mean blood melatoninvalues for childrenof comparable ages20.One of thepossible explanations for lower melatonin levels inthe population we studied is that most of our

VOLUME 12, NO.1, 1999

:,~

- -

50=.-e-

0-=

e

>Q 150r I Be....

-=Q

== 100

50

0

24 hour cycles

200 C-;0''\150

C'S I \0

1004iee 50 /2CD

U)0

66 LV ZHDANOVA ET AL.

I

patients received anti-seizure medications con-taining sodium valproate. This GABAergic com-pound has been shown to significantly suppressblood melatoninlevelsin humans21.

Increased motor activityin sleep, a conventionalsign of sleep disruption, appears to be inhibitedbythe substantial increases in serum melatonin levelsthat follow administrationof a 0.3 mg dose of thehormone. The sleep-promoting effect of the hor-mone treatment in our patients did not show adependency on either the endogenous peak mel-atonin levels or the duration of the nocturnalincrease in melatonin secretion. This result mightreflect an experimental limitation imposed by thedifferent ages and weights of the children studiedthat did not allow us to adequatelycompare them interms of endogenousmelatoninlevelsand melatoninpharmacokinetics.

Among the AS children whose melatoninsecretionwas delayed, repeated melatonintreatmentseveral hours prior to the onset of their ownnocturnal melatonin secretion resulted in a phaseadvance of their circadian rhythm. However, in thepatient who received exogenous melatoninfor morethan a month close to the time of his normalnocturnal increase in the hormone's secretion, thesleep promoting effect of the treatment was notaccompaniedby a detectable shift in the timing ofhis circadianrhythms. Thus, the repetitive, approp-riately-timed administration of melatonin to ASchildrencan phase-shifttheir circadian rhythms,butdoes not disturb the pattern if applied close to thetime of onset of endogenous nocturnal melatoninrelease.

Our data suggest that an increase in circulatingnocturnal melatonin levels within the physiologicrange, or above it, is beneficialin establishingsleepintegrity in AS children. The mechanism of theacute sleep promoting effect of melatonin remainsto be disclosed.It might be a result of a more robustendocrinesignalfor the homeostatic mechanismsofsleep initiation and sleep maintenance, a con-sequence of an acute inhibitingeffect of melatoninon the major site of the mammalianbiologicalclock- the suprachiasmaticnucleus of the hypothalamus(SCN), or anotherunknown mechanism.

Our findings in those AS children that had acircadian phase delay prior to the treatment, alsosupport the idea that melatoninhas a dual role in the

sleep/wake process: acute promotion of sleep, andthe phase shifting of an underlying circadianoscillator. This dual action of melatoninsuggests itspotential as a treatment for patientswhose insomniais related to the hormone's deficiency and/or acircadianrhythmdisorder.

Melatonin's effects in humans are poorlystudied. Since pharmacologic melatonin concen-trations in humans have been consistentlyshown todecrease body temperature22,and they are thoughtto have some effect on the development of thereproductive system23,a cautious attitude towardtreatment with the hormone is indicated and, iftreatment is initiated, a search for the minimumeffectivedose of the hormone for each individualisrecommended.

ACKNOWLEDGEMENTS

We thank Dr. H.J. Lynchfor a ftuitfuldiscussionof the data and the manuscript; the Children'sHospital nurses for their dedicated help; and Dr. A.Kartashov for helping with the statistical analysis.This work was supported in part by grants from theNational Institute of Health (IROI AG13667-01,MOl RR 02172 and MOl RR 00088-34), Centerfor Brain Sciences and Metabolism CharitableTrust, and AngelmanSyndromeFoundation.

REFERENCES

1. Clayton-Smith1. Clinical research on Angelmansyndrome in the United Kingdom: Observations of 82affected individuals. Am J Med Genet 1993; 46: 12-15.

2. Knoll JR, Nicholls RD, Lalande M. On the parentalorigin of the deletion in Angelman syndrome. HumGenet 1989; 83: 205-207.

3. Kishino T, Lalande M, Wagstaff 1. UBE3NE6-APmutations cause Angelman syndrome. Nature Genet1997; 15: 70-73.

4. Matsuura T, Sutcliffe J, Fang P, Galjaard R-J, JiangY, Benton CS, Rommens 1M, Beandet AL De novotruncating mutations in E6-AP ubiquitin-proteinligase gene (VBE3A) in Angelman syndrome. NatureGenet 1997; 15: 74-77.

5. Sheldon SR, Spire J-P, Levy HB. Sleep-wake sched-ule disorders. In: Sheldon SH, Spire J-P, Levy HB.Pediatric Sleep Medicine. Philadelphia, PA: W.B.Saunders Company, 1992; 106-111.

JOURNAL OF PEDIATRIC ENDOCRINOLOGY & METABOLISM

MELATONIN AND SLEEP IN ANGELMAN SYNDROME

6. Kennaway DJ, Stamp GE, Goble FC. Development ofmelatonin production in infants and the impact ofprematurity. J Clin Endocrinol Metab 1992; 75: 367-369.

7. Haimov I, Laudon M, Zisapel N, Souroujon M, NofD. Shlitner A. Herer P, Tzischinsky 0, Lavie P.Sleep disorders and melatonin rhythms in elderlypeople. BMl1994; 309: 167.

8. Hajak G, Rodenbeck A, Staedt J, Bandelow B,Huether G, Ruther E. Nocturnal plasma melatoninlevels in patients suffering from chronic primaryinsomnia. J Pineal Res 1995; 19: 116-122.

9. Lerner AS, Case 10. Melatonin. Fed Proc 1960; 19:590-592.

10. Anton-Tay F. Melatonin: effects on brain function.In: Costa E, Gessa GL, Sandler M, 005. Serotonin -New Vistas: Biochemistry and Behavioral andClinical Studies. New York: Raven Press, AdvBiochem Pharmaco11974; II: 234-237.

II. Cramer H, Rudolph J, Consbruch U, Kendel K. Onthe effects of melatonin on sleep and behavior inman. Adv Biochem Psychophannacol 1974; II: 187-191.

12. Dollins AB, Zhdanova IV, Wurtman RI, Lynch HJ,Deng MH. Effect of inducing nocturnal serummelatonin concentrations in daytime on sleep, mood,body temperature, and performance. Proc Natl AcadSci USA 1994; 91: 1824-1828.

13. Zhdanova IV, Wurtman RI, Morabito C, PiotrovskaVR Lynch HJ. Effects of low oral doses ofmelatonin, given 2-4 hours before habitual bedtime,on sleep in normal young humans. Sleep 1996; 19:423-431.

14. Zhdanova IV, Wurtman RI, Lynch HJ, Ives lR,Dollins AS, Morabito C, Matheson lK, Schomer DL.

VOLUME 12, NO. I, 1999

L

67

Sleep-inducing effects of low doses of melatoniningested in the evening. Clin Phann Ther 1995;57:552-558.

15. Jan IE, Espezel H, Appleton RE. The treatment ofsleep disorders with melatonin. Dev Med ChildNeuro11994; 36: 97-107.

16. McArthur AI, Budden SS. Sleep dysfunction in Rettsyndrome: a trial of exogenous melatonin treatment.Dev Med Child Neuro11998; 40: 186-192.

17. Okawa M, Takahashi K, Sasaki H. Disturbance ofcircadian rhythms in severely brain-damaged patientscorrelatedwith CT findings. J Neurol 1986;233:274-282.

18. Quine L. Sleep problems in children with mentalhandicap. J Ment Defic Res 1991; 35: 269-290.

19. Waldhauser F, Weiszenbacher G, Tatzer E. GisingerB, Waldhauser M, Schemper M, Frisch H. Alter-ations in nocturnal serum melatonin levels in humanswith growth and aging. J Clin Endocrinol Metab1988; 66: 648-652.

20. Cavallo A. Plasma melatonin rhythm in normalpuberty: interactions of age and pubertal stages.Neuroendocrinology 1992; 55: 372-379.

21. Montaleone P, Tortorella A, Borriello R. Natale M.Cassandra P, Maj M. Suppression of nocturnalplasma melatonin levels by evening administration ofsodium valproate in healthy humans. BioI Psychiatry1997; 41: 336-341.

22. Cagnacci A, Elliott JA, Yen SS. Melatonin: a majorregulator of the circadian rhythm of core temperaturein humans. J Clin Endocrinol Metab 1992: 75: ~~7-452.

23. Heubner O. Tumor der glandula pinealis. Dtsch MedWochenschr 1898; 24: 214-216.

- --