bs_bs_banner

Sleep and Biological Rhythms 2013; 11: 99–104

doi:10.1111/sbr.12010

ORIGINAL ARTICLE

Chronic sleep reduction in adolescents with Delayed Sleep Phase Disorder and effects of melatonin treatment Annette van MAANEN,1 Julia F DEWALD-KAUFMANN,1,4 Marcel G SMITS,3 Frans J OORT1,2 and Anne Marie MEIJER1 1 Research Institute of Child Development and Education, 2Department of Medical Psychology, Academic Medical Centre, University of Amsterdam, Amsterdam, 3Hospital Gelderse Vallei, Ede, The Netherlands, and 4Department of Pediatric Endocrinology, Dr. von Hauner Children’s Hospital, Ludwig- Maximilians-University, München, Germany

Abstract Homeostatic and circadian changes that occur during adolescence can result in chronic sleep reduction. This may particularly be true for adolescents with Delayed Sleep Phase Disorder (DSPD), which is associated with late Dim Light Melatonin Onset (DLMO). This study assessed the influence of melatonin treatment on chronic sleep reduction in adolescents with DSPD and examined whether adolescents with DSPD suffer from more chronic sleep reduction than adolescents from the general population before and after melatonin treatment. Adolescents with DSPD (n = 145; 55.9% boys; mean age 15,5 years; mean DLMO = 22:32 h) completed a questionnaire concerning chronic sleep reduction at baseline. From these, 53 adolescents also completed this questionnaire after on average 10 weeks of melatonin treatment. At baseline adolescents with DSPD reported significantly more symptoms of chronic sleep reduction than adolescents from the general population, whereas after treatment they reported significantly less symptoms. DLMO did not influence the effect of treatment. The improvement of chronic sleep reduction after treatment is an important finding, considering the negative consequences of chronic sleep reduction for adolescents’ daytime functioning. Key words: adolescents, chronic sleep reduction, Delayed Sleep Phase Disorder, melatonin.

INTRODUCTION According to the two-process model, sleep is regulated by a homeostatic process (process S) in which sleep pressure increases with longer wake time and a circadian process (process C) that determines periods during the day when one is most likely to fall asleep. The interplay between these two processes regulates the period and time of sleep.1 As described by Carskadon et al.,2 process S and process C undergo several changes during adolescence Correspondence: Ms Annette van Maanen, University of Amsterdam, Nieuwe Prinsengracht 130, 1018 VZ Amsterdam, the Netherlands. Email: [email protected] Accepted 15 January 2013.

that elicit later sleep times. Due to phase delay related to pubertal development, slower accumulation of homeostatic sleep pressure and environmental factors, adolescents’ sleep is shifted towards a later time. As adolescents often have to get up early in the morning to go to school, many adolescents do not obtain enough sleep.3 Insufficient sleep is often experienced for a longer period of time, which can result in chronic sleep reduction. Chronic sleep reduction is defined as insufficient and/or poor sleep over a relatively long time period,4 and can have severe negative daytime consequences, such as impaired academic performance, depressive symptoms, and increased risk-taking.5 One particular group that might suffer from chronic sleep reduction is the group of adolescents with Delayed Sleep Phase Disorder (DSPD). DSPD is a

© 2013 The Authors Sleep and Biological Rhythms © 2013 Japanese Society of Sleep Research

99

A van Maanen et al.

disorder characterized by a delay of the sleep period (indicated by late melatonin onset)6,7 in relation to the desired time of sleep, resulting in difficulties falling asleep or waking at the desired time,8 and often extreme sleepiness during the (school) day.3 The exact etiology of DSPD is not known, but it is assumed to result from abnormalities in process C or deficits in process S. In adolescents with DSPD, the homeostatic and circadian changes that occur during adolescence are considered to be extreme.3 Exogenous melatonin, if applied at the right time and in the right dose, can phase shift the circadian rhythm and advance sleep onset,9,10 and as a consequence may diminish symptoms of chronic sleep reduction (shortness of sleep, loss of energy, sleepiness, irritation). Previous research has shown that melatonin treatment resulted in a decrease of sleepiness in a sample of adolescents with DSPD.11 However, another study in adults did not find an effect of melatonin on sleepiness.12 The present study investigates the effects of melatonin treatment on chronic sleep reduction in adolescents with DSPD. We also examined whether baseline DLMO influences the effects of treatment, as there are indications from previous studies that this might be the case.13,14 In addition, pre- and post-treatment scores are compared with a control group from the general population.

METHODS Participants Participants were 145 adolescents (55.9% boys, 10.7– 20.1 years, mean age 15.5 ⫾ 1.9 years) with delayed sleep phase (mean self-reported bedtime 22:26 ⫾ 1:19 h, mean self-reported sleep onset latency 124 ⫾ 68 min) and late Dim Light Melatonin Onset (DLMO) (mean DLMO 22:32 ⫾ 1:05 h). They were referred to the hospital by their general practitioner because of their sleep onset problems. Inclusion criteria for participation were: (i) age between 10 and 20 years old (similar to the review of Crowley et al.);3 (ii) chronic sleep onset problems defined as (a) complaints of sleep-onset problems expressed by parents and/or child, (b) occurrence on at least 4 days/week for longer than 1 year, (c) average sleep onset later than 20:30 hours for 10 year old children and for older children 15 min later per year, and (d) average sleep latency exceeding 30 min; and (iii) DLMO later than 20:00 hours. Excluded were children who used melatonin before inclusion.

100

The control group consisted of 166 adolescents from the general population recruited from high schools in and around Amsterdam for an earlier study (28% boys, 12.2–16.5 years, mean age 15.2 ⫾ 0.6 years).15 All participants were attending years 3 or 4 of the highest high school level in the Netherlands. No further inclusion or exclusion criteria were specified.

Procedure Before adolescents with DSPD had their first appointment in the hospital, DLMO was measured in saliva. Adolescents had to chew on cotton plugs hourly according to a predetermined schedule for one evening. They were instructed to do this in a dim light environment, as bright light reduces melatonin secretion. All adolescents completed the Chronic Sleep Reduction Questionnaire (CSRQ)4 prior to treatment. Participants with DSPD received melatonin treatment in the Centre for Sleep–Wake Disorders and Chronobiology of Hospital Gelderse Vallei in Ede, the Netherlands. The dose was determined for each adolescent individually. Treatment was started with 1 mg. If sleep did not change within 1 week, the dose was increased every week with 1 mg until changes in sleep onset occurred. If 1 mg melatonin was effective the dose was lowered until the lowest effective dose was reached. This was done to prevent loss of response due to slow melatonin metabolization.16 The lowest prescribed dose was 0.1 mg and the highest dose was 10 mg. The mean dose was 2.7 mg. The timing of administration was also determined individually, 3 h before individual DLMO, however not earlier than 19:00 h. The majority of adolescents (88.2%) took melatonin at 19:00, 19:30 or 20:00 h. Of the 145 adolescents with DSPD, 53 adolescents also completed the CSRQ after melatonin treatment. To check the randomness of missing, we tested for baseline differences with the 92 adolescents who completed the first measurement only, but found no significant differences on CSRQ, DLMO, and gender. Still, the group missing at the second measurement occasion was found to be 0.7 years older, so we included age as a covariate in our analyses. The average time between the first and second measurement occasion was 10 weeks. The measurements were spread over the year (i.e. in all seasons).

Measures The Chronic Sleep Reduction Questionnaire (CSRQ)4 is a 20-item questionnaire consisting of four subscales that

© 2013 The Authors Sleep and Biological Rhythms © 2013 Japanese Society of Sleep Research

Chronic sleep reduction and melatonin

Table 1 Mean scores of adolescents with DSPD before and after melatonin treatment and of adolescents from the general population Comparison of DSPD group before treatment (n = 145) with general population sample (n = 166)

Comparison before and after treatment for DSPD group (n = 53)

CSRQ total scale Shortness of sleep Sleepiness Loss of energy Irritation

Before treatment

After treatment

Effect size†

DSPD group

General population sample

Effect size†

Mean (SD)

Mean (SD)

D

Mean (SD)

Mean (SD)

d

40.88 12.77 8.33 11.41 8.37

(6.93) (1.69) (2.46) (2.84) (2.71)

28.88 7.96 5.66 7.65 7.61

(8.08) (2.81) (2.09) (2.66) (2.91)

-1.58 -1.96 -1.16 -1.37 -0.27

41.76 12.98 8.39 11.54 8.86

(6.93) (1.98) (2.34) (2.75) (2.88)

32.35 10.26 6.73 8.19 7.17

(6.48) (2.47) (1.90) (2.25) (1.91)

1.41 1.21 0.78 1.34 0.70

†

Effect sizes d of 0.2, 0.5, and 0.8 are considered small, medium, and large.17

measure different symptoms of chronic sleep reduction: “Shortness of sleep” (6 items; e.g. “I am a person who does not get enough sleep”), “Sleepiness” (4 items; e.g. “Do you feel sleepy during the day?”), “Loss of energy” (5 items; e.g. “I am active during the day”), and “Irritation” (5 items; e.g. “Others think that I am easily irritated”). The sum of the scales yields a total score that can vary between 20 and 60. A higher score indicates more symptoms of chronic sleep reduction. One item (item 7: “At noon I feel as energetic as in the morning”) was excluded from the analyses, because for the majority of the adolescents with DSPD, none of the response categories was applicable. The CSRQ has good reliability and was recently validated against actigraphy data.15 In the present study Cronbach’s alpha of the total scale was 0.84 at baseline and 0.89 after treatment, and for the subscales 0.63 and 0.75 (Shortness of sleep), 0.76 and 0.76 (Sleepiness), 0.81 and 0.82 (Loss of energy), and 0.84 and 0.85 (Irritation) for baseline and posttreatment measures respectively.

Analyses To investigate whether chronic sleep reduction diminishes after treatment, repeated measures analysis of variance was applied, with age and gender and interaction effects of age and gender with treatment as covariates. To investigate whether effectiveness of melatonin depends on DLMO, we conducted a separate analysis with DLMO and age as covariates. To gain more insight into the effect of DLMO, means and standard deviations were calculated for adolescents with a DLMO earlier than the mean and adolescents with a DLMO later than the mean. Comparison with

mean scores of the control group from the general population was done by conducting t-tests for independent samples.

RESULTS In comparison to baseline, chronic sleep reduction significantly diminished after melatonin treatment: adolescents improved on all CSRQ subscales and effect sizes were large (see Table 1). With age, gender, and their interactions with treatment as covariates, and the total CSRQ scale and subscales Shortness of sleep, Sleepiness, Loss of energy, and Irritation separately as dependent variables, we found a significant interaction effect on Sleepiness (see Fig. 1): the positive effect of age on Sleepiness at baseline (B = 0.90, se = 0.39, P = 0.03) disappeared after treatment (B = 0.09, se = 0.35, P = 0.80). With covariates, there was no treatment effect on Irritation. We found small positive effects of age on the total score and on Shortness of Sleep, Sleepiness, and Loss of Energy, but these were not significant at the 5% level. See Table 2 for full results. Next, we investigated whether DLMO influences the effect of melatonin treatment. Adolescents with a DLMO earlier than the mean showed significant lower scores on chronic sleep reduction than adolescents with a DLMO later than the mean on both baseline and after treatment, except on the scale Loss of energy at baseline, and on Irritation at baseline and after treatment (Table 3). However, in the models that included baseline DLMO, age and their interactions as covariates, only main effects were found for treatment and no significant interaction effects with treatment were found. The absence of an interaction effect with treatment

© 2013 The Authors Sleep and Biological Rhythms © 2013 Japanese Society of Sleep Research

101

A van Maanen et al.

age < 15.5 years > 15.5 years

Sleepiness

9,00

8,00

7,00

6,00

baseline

after treatment

Treatment

Figure 1 Comparison of treatment effects on sleepiness of children in early and late adolescence.

suggests that the effect of treatment was not influenced by DLMO. Finally, we compared the scores of the participants with DSPD with a control group from the general population. Adolescents with DSPD had a mean score of 41.76 on the CSRQ at baseline. This was significantly higher (t = 12.37, d.f. = 309, P < 0.01) than the mean score that was found in the group of adolescents from the general population (mean score = 32.35 (a small discrepancy with the Dewald et al. paper stems from removing one item from the 20-item CSRQ, as it did not apply to the DSPD sample).15 Strikingly, after melatonin treatment, adolescents with DSPD scored even lower than the adolescents from the general population (e.g. total score averages of 28.88 and 32.35 yield t = -3.19, d.f. = 217, P < 0.01, and effect size d = -0.50).

DISCUSSION This study found that adolescents with DSPD show more symptoms of chronic sleep reduction than adolescents from the general population. Melatonin treatment significantly decreased chronic sleep reduction to a level that was even lower than that of adolescents from the general population. The decrease in chronic sleep reduction we found after melatonin treatment is in accordance with earlier research reporting a reduction in sleepiness in adolescents treated with melatonin,11 but is in contrast with research conducted in adults that did not find an effect

102

on sleepiness.12 However, in the latter study sleepiness measures were completed on the second night of the treatment phase, whereas in our study measurements were taken after several weeks of treatment. This might explain the conflicting results. Although adolescents with DLMO earlier than the mean and later than the mean scored differently on the CSRQ subscales, their decrease in symptoms of chronic sleep reduction was not significantly different, which suggests that treatment effects were not influenced by baseline DLMO. This can be considered a positive result, as it indicates that melatonin treatment works for all DSPD children, regardless of their DLMO. However, one should keep in mind that the time of melatonin administration was defined with reference to DLMO. Moreover, there are no guidelines on what the time of DLMO should be in adolescents with normal sleep with reference to their age and, as a consequence, there are no established criteria for determining which deviations from normal DLMO according to age, constitute DSPD. Therefore, the mean DLMO reported in our study (22:32 h) can be different from DLMO reported in other studies10 and should be evaluated with reference to the age groups in question. In a previous study, we found that a reduction in behavior problems in children after melatonin treatment was stronger for children with an earlier DLMO.13 Still, another study found that treatment effects on sleep onset and DLMO were stronger for children who had a later baseline DLMO.14 In view of these inconclusive results and the limited sample size in our study, the influence of DLMO on melatonin treatment should be further investigated. The finding that chronic sleep reduction after melatonin treatment was even lower than in the general population, can be explained by the fact that the general population sample may also include adolescents with sleeping problems, whereas adolescents with DSPD are good sleepers after melatonin treatment. Also, adolescents with DSPD might have the tendency to give more positive answers after treatment, as they compare their sleep with their sleep before treatment. In addition, expectation effects may play a role, meaning that adolescents with DSPD report a strong reduction of symptoms because they had high expectations of treatment effects.18 As discussed by Crowley et al.,3 the changes that occur in process S and process C during adolescence might be extreme in DSPD. From the literature it is clear that adequate treatment with melatonin changes circadian rhythmicity.9,10 Apparently, changing process C

© 2013 The Authors Sleep and Biological Rhythms © 2013 Japanese Society of Sleep Research

Chronic sleep reduction and melatonin

Table 2 Results of repeated measures analysis of variance comparing the CSRQ-scores at baseline and after melatonin treatment for adolescents with DSPD Models with only treatment as predictor

CSRQ total scale

Shortness of sleep

Sleepiness

Loss of energy

Irritation

Treatment Gender Age Treatment*gender Treatment*age Treatment Gender Age Treatment*gender Treatment*age Treatment Gender Age Treatment*gender Treatment*age Treatment Gender Age Treatment*gender Treatment*age Treatment Gender Age Treatment*gender Treatment*age

Models with covariates

F (df)

P

149.92 (1,52)

<0.01

208.08 (1,52)

<0.01

77.42 (1,52)

<0.01

90.38 (1,52)

<0.01

3.99 (1,52)

0.05

F (df) 11.71 0.50 2.59 0.33 1.52 13.94 0.97 2.68 0.61 0.71 7.05 0.36 2.26 0.25 5.77 11.46 0.27 2.06 0.04 2.95 0.01 3.10 0.18 0.38 0.15

P

(1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49) (1,49)

<0.01 0.48 0.11 0.57 0.22 <0.01 0.33 0.11 0.44 0.40 0.01 0.55 0.14 0.62 0.02 <0.01 0.61 0.16 0.84 0.09 0.92 0.08 0.67 0.54 0.70

Note: results are shown for models with only treatment as predictor and for models with age (standardized), gender and interaction effects of age and gender with treatment added as covariates.

Table 3 Mean scores of adolescents with DSPD with DLMO earlier or later than the mean† Baseline comparison

CSRQ total scale Shortness of sleep Sleepiness Loss of energy Irritation

After treatment comparison

DLMO earlier than the mean (n = 60)

DLMO later than the mean (n = 45)

Effect size‡

DLMO earlier than the mean (n = 24)

DLMO later than the mean (n = 19)

Effect size‡

Mean (SD)

Mean (SD)

d

Mean (SD)

Mean (SD)

d

40.31 12.45 7.85 11.25 8.76

(7.87) (2.06) (2.38) (2.89) (3.02)

43.53 13.44 9.05 12.12 8.91

(6.06) (1.67) (2.11) (2.53) (2.82)

-0.45 -0.52 -0.53 -0.32 -0.05

26.13 7.21 4.92 7.08 6.92

(5.75) (2.15) (1.02) (2.26) (2.43)

32.54 8.84 6.47 8.75 8.47

(9.32) (3.06) (2.74) (3.13) (3.37)

-0.85 -0.63 -0.79 -0.62 -0.54

† DLMO was only available for 105 adolescents. Most of the missings were caused by uninterpretable melatonin curves. ‡Effect sizes d of 0.2, 0.5, and 0.8 are considered small, medium, and large.17

© 2013 The Authors Sleep and Biological Rhythms © 2013 Japanese Society of Sleep Research

103

A van Maanen et al.

indeed results in a decrease in symptoms of chronic sleep reduction. Interestingly, at baseline 52 out of 53 adolescents reported that they wanted to sleep longer (CSRQ item 18), but after treatment only 11 (20.8%). This suggests that advancing sleep onset by changing process C might also affect sleep pressure (as expressed by the wish to sleep longer). Before treatment this high sleep pressure appears not to be sufficient for inducing sleep in case the circadian process is disturbed. This raises a question about the role of process S in adolescents with DSPD and how this is influenced by melatonin. In this respect, it has to be considered that besides the chronobiotic (phase shifting) effect, melatonin also has a soporific (sleep inducing) effect.19 Thus, it not only affects the circadian system, but it might also affect (accumulation of) sleep pressure. A limitation of this study is that no information is available concerning sleep improvement or phase-shift advance after treatment. For this reason, we do not know whether there are differences in CSRQ scores between good responders and non-responders to melatonin treatment. Although conclusions about causality cannot be drawn due to the lack of a randomized placebo group, comparison with a normative control group clearly indicates that melatonin treatment diminishes symptoms of chronic sleep reduction in such a way that adolescents with DSPD have even less symptoms than reported by adolescents from the general population. Moreover, our results indicate that all children equally benefit from melatonin treatment, independent of their DLMO. These are important findings, considering the negative consequences of chronic sleep reduction for adolescents’ daytime functioning.

REFERENCES 1 Borbély AA. A two process model of sleep regulation. Hum. Neurobiol. 1982; 1: 195–204. 2 Carskadon MA, Acebo C, Jenni OG. Regulation of adolescent sleep: implications for behavior. Ann. N. Y. Acad. Sci. 2004; 1021: 276–91. 3 Crowley SJ, Acebo C, Carskadon MA. Sleep, circadian rhythms, and delayed phase in adolescence. Sleep Med. 2007; 8: 602–12. 4 Meijer AM. Chronic sleep reduction, functioning at school and school achievement in preadolescents. J. Sleep Res. 2008; 17: 395–405.

104

5 Moore M, Meltzer LJ. The sleepy adolescent: causes and consequences of sleepiness in teens. Paediatr. Respir. Rev. 2008; 9: 114–21. 6 Rahman SA, Kayumov L, Tchmoutina EA, Shapiro CM. Clinical efficacy of dim light melatonin onset testing in diagnosing delayed sleep phase syndrome. Sleep Med. 2009; 10: 549–55. 7 Wyatt JK, Stepanski EJ, Kirkby J. Circadian phase in Delayed Sleep Phase Syndrome: predictors and temporal stability across multiple assessments. Sleep 2006; 29: 1075–80. 8 American Academy of Sleep Medicine. ICSD - International classification of sleep disorders, revised: Diagnostic and coding manual. American Academy of Sleep Medicine, 2001. 9 Sack RL, Auckley D, Auger R et al. Circadian rhythm sleep disorders: part II, advanced sleep phase disorder, delayed sleep phase disorder, free-running disorder, and irregular sleep-wake rhythm. Sleep 2007; 30: 1484–501. 10 Van Geijlswijk IM, Korzilius HPLM, Smits MG. The use of exogenous melatonin in Delayed Sleep Phase Disorder: a meta-analysis. Sleep 2010; 33: 1605–14. 11 Szeinberg A, Borodkin K, Dagan Y. Melatonin treatment in adolescents with Delayed Sleep Phase Syndrome. Clin. Pediatr. 2006; 45: 809–18. 12 Kayumov L, Brown G, Jindal R, Buttoo K, Shapiro CM. A randomized, double- blind, placebo-controlled crossover study of the effect of exogenous melatonin on Delayed Sleep Phase Syndrome. Psychosom. Med. 2001; 63: 40–8. 13 Van Maanen A, Meijer AM, Smits MG, Oort FJ. Melatonin and sleep effects on health, behavior problems and parenting stress. Sleep Biol. Rhythms 2011; 9: 165–71. 14 Van der Heijden KB, Smits MG, van Someren EJW, Gunning WB. Prediction of melatonin efficacy by pretreatment dim light melatonin onset in children with idiopathic chronic sleep onset insomnia. J. Sleep Res. 2005; 14: 187–94. 15 Dewald JF, Short MA, Gradisar M, Oort FJ, Meijer AM. The Chronic Sleep Reduction Questionnaire (CSRQ): a cross-cultural comparison and validation in Dutch and Australian adolescents. J. Sleep Res. 2012; 21: 584–94. 16 Braam W, Van Geijlswijk I, Keijzer H et al. Loss of response to melatonin treatment is associated with slow melatonin metabolism. J. Intellect. Disabil. Res. 2010; 54: 547–55. 17 Cohen J. Statistical Power Analysis for the Behavioral Sciences, Second edn. Erlbaum: Hillsdale, NJ, 1988. 18 Turner JA, Deyo RA, Loeser JD, Von Korff M, Fordyce WE. The importance of placebo effects in pain treatment and research. JAMA 1994; 271: 1609–14. 19 Arendt J, Skene DJ. Melatonin as a chronobiotic. Sleep Med. Rev. 2005; 9: 25–39.

© 2013 The Authors Sleep and Biological Rhythms © 2013 Japanese Society of Sleep Research