What predicts inattention in adolescents? An experience-sampling study comparing chronotype, subjective, and objective sleep parameters

Accepted Manuscript What predicts inattention in adolescents? An experience-sampling study comparing chronotype, subjective and objective sleep parameters Timo Hennig, Katarina Krkovic, Tania M. Lincoln PII:

S1389-9457(17)30302-7

DOI:

10.1016/j.sleep.2017.07.009

Reference:

SLEEP 3457

To appear in:

Sleep Medicine

Received Date: 21 April 2017 Revised Date:

22 June 2017

Accepted Date: 9 July 2017

Please cite this article as: Hennig T, Krkovic K, Lincoln TM, What predicts inattention in adolescents? An experience-sampling study comparing chronotype, subjective and objective sleep parameters, Sleep Medicine (2017), doi: 10.1016/j.sleep.2017.07.009. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Running head: CHRONOTYPE, SLEEP, AND INATTENTION

1

RI PT

What predicts inattention in adolescents? An experience-sampling study comparing chronotype,

SC

subjective and objective sleep parameters

Timo Hennig *, Katarina Krkovic, Tania M. Lincoln

M AN U

Clinical Psychology and Psychotherapy, Institute of Psychology, University of Hamburg, Hamburg, Germany

* Corresponding author., Universität Hamburg, Clinical Psychology and Psychotherapy, Institute

TE D

of Psychology, Faculty of Psychology and Human Movement Science, Von-Melle-Park 5, 20146 Hamburg, Germany. Tel.: +49 40 42838 7820; fax: +49 40 42838 6170,

EP

E-mail address: [email protected] (T. Hennig).

Abstract

AC C

Objective: Many adolescents sleep insufficiently, which may negatively affect their

functioning during the day. To improve sleep interventions, we need a better understanding of the specific sleep-related parameters that predict poor functioning. We investigated to which extent subjective and objective parameters of sleep in the preceding night (state parameters) and the trait variable chronotype predict daytime inattention as an indicator of poor functioning.

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

2

Methods: We conducted an experience-sampling study over 1 week with 61 adolescents (30 girls, 31 boys, mean = 15.5 years, standard deviation (SD) = 1.1). Participants rated their inattention two times each day (morning, afternoon) on a smartphone. Subjective sleep

RI PT

parameters (feeling rested, positive effect upon awakening) were assessed each morning on the smartphone. Objective sleep parameters (total sleep time, wake after sleep onset, sleep

efficiency) were assessed with a permanently worn actigraph. Chronotype was assessed with a

SC

self-rated questionnaire at baseline. We tested the effect of subjective and objective state

parameters of sleep on daytime inattention, using multilevel multiple regressions. Then, we

M AN U

tested whether the putative effect of the trait parameter chronotype on inattention is mediated through state sleep parameters, again using multilevel regressions.

Results: We found that short sleep time, but no other state sleep parameter, predicted inattention to a small effect. As expected, the trait parameter chronotype also predicted

TE D

inattention: morningness was associated with less inattention. However, this association was not mediated by state sleep parameters.

Conclusions: Our results indicate that short sleep time causes inattention in adolescents.

EP

Extended sleep time might thus alleviate inattention to some extent. However, it cannot alleviate the effect of being an ‘owl’.

AC C

Keywords:

Circadian preference Morningness Eveningness Daily diary

Ambulatory assessment

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

3

Introduction Sufficient and good sleep is an important prerequisite for being attentive and functioning well on the following day. Many adolescents, however, have poor sleep routines leading to an

RI PT

insufficient amount of sleep [1,2]. This is probably due to the fact that adolescents experience bioregulatory pressure towards delayed sleeping caused by maturationally driven changes in the circadian timing system and slowed increase of sleep pressure [3]. The preference for delayed

SC

sleep onset is further corroborated by psychological factors such as late screen time and social networking [4]. On the other hand, there is a societal pressure to rise early, mainly due to the

M AN U

early starting times at school [5]. Thus, many adolescents do not sleep enough, which has been found to be associated with poor performance in a variety of domains, such as academic achievement [6–9], attention and executive functioning [10], working memory performance [11], memory consolidation [12] as well as with risky behavior [13] and mental health problems [9,14].

TE D

There is evidence that not only objective but also subjective sleep parameters predict poor functioning or are even more important. Using multiple regression, Boschloo et al. [15] found adolescent-reported sleepiness in the morning but not time in bed to predict poor grades

EP

and school performance. Similarly, van der Heijden et al. [16] found feeling rested upon awakening, but not sleep duration, to be associated with reaction time and behavioral and

AC C

attention problems in school children. In this context, it is important to keep in mind that there might be underlying traits that

account for both insufficient sleep and poor daytime functioning. A likely candidate for such a trait is the chronotype that refers to preferred sleep and wake times and day times of peak performance [17]. Compared to morning types or ‘larks’, evening types or ‘owls’, who experience their daily ‘high’ later in the day, have been found to be at a higher risk for poor

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

4

school performance [8,18,19], poor mental health and behavior [20,21], and lower functioning [22]. The chronotype is thought to be a trait with genetic factors determining up to half of its variance [23]. Escribano et al. [24] found both adolescent-reported sleep time and evening

RI PT

chronotype to be associated with poor school performance. Van der Heijden et al. [16] found that the evening chronotype was associated with behavioral and attention problems and that this association was mediated by feeling unrested upon awakening. However, these findings are

SC

based on cross-sectional studies with subjective reports of sleep and need to be replicated in experience-sampling or longitudinal studies with objective sleep assessments such as actigraphy.

M AN U

If we could assess more reliably what causes inattention during the day, for example, whether it is the short sleep time or the appraisal of one’s sleep as being more or less restful, we could use this knowledge to improve sleep interventions.

In this study, we thus assessed subjective and objective sleep parameters and self-rated

TE D

inattention in an experience-sampling design combined with actigraphy over a period of 1 week. The experience-sampling design has the advantage that temporal sequences in the occurrence of symptoms can be examined, which is a prerequisite of causality [25]. Actigraphy has the

EP

advantage that it enables an objective, non-invasive, ecologically valid assessment of sleep in adolescents [26,27].

AC C

First, we expected both subjective and objective sleep parameters to be associated with

inattention. Second, to determine which sleep parameter primarily predict inattention, we tested all parameters that correlated with inattention in a multiple regression analysis. Third, we expected that the chronotype would influence both state sleep parameters and inattention. Finally, we tested whether the putative effect of chronotype on inattention is mediated through those state sleep parameters that predicted inattention.

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

5

Methods Sample We recruited a community sample of adolescents aged 14–17 years through

RI PT

announcements in public places (eg, shops and sports centers) and on the university web page. Participants received monetary compensation after study completion. All participants and their parents gave full informed consent. The study was approved by the ethics committee of the

SC

German Psychological Society (reference number TH_022015). Schedule of Assessments

M AN U

The schedule of the assessments is shown in Fig. 1. The study period included the night after a baseline assessment at the university and the 7 subsequent days. At baseline, participants were handed an actigraph and a smartphone for the entire study period. Participants were instructed to permanently wear the actigraph and to complete three self-report assessments per

TE D

day on the smartphone (upon awakening/before school, at noon/after school, in the evening). Participants could manually start the awakening assessments after 05:00 h, the noon assessments after 11:30 h, and the evening assessments after 18:00 h. If assessments had not

EP

been started manually or were incomplete, the smartphones were programed to ring and to remind the subject to complete the assessment (all other smartphone services were disabled). In

AC C

order to respect individual routines (eg, waking times, class schedules), a study collaborator consulted with each participant at baseline about when to set the exact reminder times for each day of the assessment period. Each assessment could be postponed up to 1 h after the initial reminder, thereafter the according assessment was defined as missing.

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

6

Instruments Chronotype At baseline, participants completed the German version of the reduced Morningness-

RI PT

Eveningness Questionnaire rMEQ [28]. The rMEQ consists of five self-rated items on circadian preference (eg, “Considering only your own ‘feeling best’ rhythm, at what time would you get up if you were entirely free to plan your day?”, “At what time in the evening do you feel tired

SC

and as a result in need of sleep?”). The item ratings are transformed and summed up to an overall score ranging from 4 to 25, with higher scores indicating greater morningness. Cut-off scores are

M AN U

used to classify the chronotype (4–11 evening type, 12–17 neither/intermediate type, and 18–25 morning type [28]). The rMEQ is a standard instrument to assess circadian preference in adults [28] and has also been used in adolescent samples [29–31]. The internal consistency in the present sample was acceptable with α = 0.70.

TE D

Subjective Sleep Measures

Upon awakening, participants rated the quality of sleep during the preceding night on a Motorola Moto G smartphone using the movisensXS experience-sampling software for Android

EP

(version 4.4). We used the item “How restful was your sleep?” (to be answered on a five-point scale from “not at all” to “very”) from the sleep diary of the German Sleep Society [32] as a

AC C

direct measure of sleep quality. Moreover, participants rated their positive effect (PA) upon awakening with three items as an indirect measure of sleep quality. Each PA item represented the positive pole of one of three mood dimensions based on the state mood assessment by Wilhelm and Schoebi [33]. The items started with “This morning, I felt…” followed by “happy, content, in a good mood” (valence), “relaxed, calm, balanced” (calmness), and “full of energy, motivated, attentive” (energetic arousal). Each item was presented with a blank scale below it with the

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

7

ending labels “not at all” and “very strong”. Participants had to click on the scale for a movable slider to appear. For the purpose of the analyses, but not visible to the participants, the slider position was translated into scores between 0 and 100. Multilevel internal consistency was

We used the mean score of the three items for the further analyses. Objective Sleep Measures

RI PT

satisfactory with α = 0.82 on the within-subjects level and α = 0.94 on the between-subjects level.

SC

We used actigraphy to assess participants’ sleep as it is a non-invasive, ecologically valid assessment, recommended for use in pediatric sleep research [26,27]. Participants carried an

M AN U

Actiwatch 2 (Philips Respironics) on their non-dominant wrist over the entire study period. Data was collected at 30-s epochs which is one of the most common and validated settings [34]. We used the Actiware software (version 6.0.4) with default settings (medium threshold for wake phase identification, 10 inactive minutes to define sleep onset and offset) to calculate three

TE D

objective sleep parameters: total sleep time defined as the time sleeping in hours without nocturnal wake intervals; sleep efficiency defined as the percentage of time sleeping in relation to time in bed; and wake after sleep onset defined as the sum of nocturnal wake intervals in

EP

minutes. The Actiwatch 2 has been found to have good sensitivity to correctly identify sleep in adolescents (0.97), however, poorer specificity to correctly identify wake time (0.54), resulting

AC C

in an overall satisfactory accuracy (0.88) [35]. Inattention

Participants rated their inattention two times each day on the smartphone. Each day at

noon, participants rated their inattention during the morning. Each evening, participants rated their inattention during the afternoon. We used three items to assess inattention based on the DSM-5 definition of Attention/Deficit-Hyperactivity Disorder [36]. The noon items started with

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

8

“This morning, …”, followed by “I had difficulties to get things done”, “I was easily distracted and could not concentrate over a longer period of time”, “my mind shifted or I had daydreams”. The evening items were identical to the noon items but started with “This afternoon, …”. Each

RI PT

item was presented with a blank scale below it with the ending labels “not at all” and “all the time”. Participants had to click on the scale for a movable slider to appear whose position was translated into scores between 0 and 100. Multilevel internal consistency was satisfactory with α

mean score of the three items for the further analyses.

M AN U

Analyses

SC

= 0.67 on the within-subjects level and α = 0.79 on the between-subjects level. We used the

First, we tested the correlations between the sleep parameters and inattention at the within-subjects and between-subjects level using the function ‘StatsBy’ of the R package ‘psych’. We calculated correlations for both the chronotype groups and for the continuous variable

TE D

morningness. Second, we conducted a multiple multilevel regression (unrestricted and controlled for weekday vs weekend, and morning vs afternoon inattention), using the SPSS 24 linear mixed-models procedure with random intercepts and fixed slopes. Thereby, the single data-

EP

points (within-subjects level) were nested within participants (between-subjects level). We used those sleep parameters that significantly correlated with inattention as predictors and inattention

AC C

as the outcome variable. The sleep parameters were entered person-mean centered to the analysis to enable a direct interpretation of the results [37]. Third, we examined the direct effect of the chronotype (using the continuous morningness variable) on inattention and the indirect effect through those sleep parameters that significantly predicted inattention. Each path was tested with a multilevel regression, following established approaches to mediation [38].

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

9

The variables on positive effect, wake after sleep onset, sleep efficiency, and inattention were skewed. To account for non-normality, we used nonparametric bootstrapping with 10,000

Results

RI PT

replications and bias-corrected accelerated confidence intervals.

The sample included 30 girls and 31 boys with a mean age of 15.1 years (SD = 1.1). The assessments took place at 272 week-/schooldays and 142 weekend/free days. Table 1 shows the

SC

distribution of the sleep parameters, inattention, and assessment times. According to the cutoff criteria, 38 participants fulfilled criteria for the evening chronotype, 23 participants were

M AN U

intermediate type, and no participant fulfilled criteria for morning type. Evening chronotypes completed the awakening assessments 22 min later than intermediate types (t = 3.29, p<0.01), the noon assessment 8 min later (t = 1.26, p=0.21), and the evening assessment 31 min later (t = 2.64, p<0.01). Complete data were available for 761 data points (10.9% missing).

TE D

The intraclass correlation (ICC) for inattention was 0.30, indicating that 30% of the variance was due to differences between participants and 70% to differences within participants (over the days). Participants with evening chronotype reported more inattention (mean = 27.08,

p<0.01.

EP

SD = 20.26) than those who were intermediate type (mean = 20.81, SD = 19.74): t = 4.32,

AC C

The correlations of the sleep parameters and inattention are shown in Table 2. Inattention

was negatively correlated with all sleep variables except sleep efficiency at the within-subjects level. Thus, we entered all state sleep variables except sleep efficiency into the regression analysis as potential predictors of inattention. Inattention was negatively correlated with chronotype (evening vs intermediate type) but not with the continuous morningness variable at the between-subjects level.

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

10

The results of the multiple multilevel regression are shown in Table 3. When we analyzed all state sleep variables simultaneously in the unrestricted model, total sleep time was the only predictor of inattention. The effect was small: 1 h more sleep predicted 1.52 units less

RI PT

inattention. As can be seen from the controlled model, the day type predicted inattention with less inattention on weekends. Total sleep time was the only other significant predictor in the controlled model; however, the effect was smaller than in the unrestricted model and almost

SC

insignificant. The time of day of the inattention assessment did not predict inattention.

Figure 2 shows the investigated mediation from chronotype (ie, morningness) on

M AN U

inattention through total sleep time. All paths were significant: morningness predicted inattention and total sleep time; total sleep time predicted inattention. When we controlled for day type, the effects were smaller but remained significant. When we analyzed morningness and total sleep time simultaneously as predictors of inattention, there was no reduction in the regression weights,

TE D

indicating that there was no mediation.

Discussion

In this study, we investigated which parameters of sleep predict daytime inattention, as an

EP

indicator of functioning and performance, in adolescents, using an experience-sampling and actigraphy design. Analyzed separately, the feeling rested and positive effect upon awakening,

AC C

total sleep time and wake after sleep onset were all associated with inattention. Analyzed simultaneously, however, total sleep time was the only parameter that significantly predicted inattention. In contrast to previous studies [15,16], we did not find subjective parameters to predict inattention.

In line with the literature [19,22], we found the chronotype to be associated with daytime inattention, with adolescents of the evening type being more inattentive than adolescents of

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

11

intermediate type. To determine how the trait chronotype is linked with current sleep, we investigated the potential mediation of the effect of chronotype on inattention through total sleep time, similar to the cross-sectional study by van der Heijden et al. [16]. However, although

RI PT

chronotype was associated with sleep time, we did not find evidence of a mediation. This

indicates that chronotype and sleep time exert an independent effect on inattention, which is in line with the findings by Escribano et al. [24].

SC

On the within-subjects level, the correlation between wake after sleep onset and

inattention was significantly negative, which is counterintuitive. However, when we controlled

M AN U

for total sleep time (not depicted) that was strongly correlated with both other variables, the correlation between wake after sleep onset and inattention disappeared, indicating a confounder effect of total sleep time.

Before discussing the implications, we would like to point to some limitations of the

TE D

study. One is that we exclusively relied on self-ratings of inattention as an outcome variable. Our findings are thus restricted to a subjective evaluation of inattention which may be biased. Future studies should include additional assessments of inattention, for example, teacher ratings or

EP

neurocognitive testing. However, adolescent self-reported inattention is nevertheless a significant outcome because it has been found to correlate with poor planning skills and executive

AC C

functioning [39], depression, anxiety, and conduct problems, and to predict poor peer relations [40]. Another limitation is that we investigated a convenience sample that might not be fully representative of the age group investigated. A limitation concerning actigraphy is that its sensitivity to detect wakefulness during sleep periods has been questioned [27,35]. Thus, we cannot be certain about the validity of our estimations of sleep efficiency and wake after sleep onset – parameters we found no effects for. As a validation, the actigraphy data could be

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

12

analyzed with software settings other than default (medium threshold), especially, using the lowthreshold setting that also has good psychometric properties [35]. An objective validation of the self-reported chronotype could have been the (corrected) midpoint of sleep on free days (ie,

RI PT

without social obligations) as derived from the actigraphy data [41,42]. However, even though we have information on day type available, we cannot verify that the weekend was free of social obligations given that many adolescents have plenty of weekend activities such as church, sports,

SC

family, and friends. Finally, even though our experience-sampling design has the major

advantage that temporal sequences can be examined [25], temporal precedence is not sufficient

M AN U

to prove causality. Subsequent studies should directly aim to manipulate sleep and assess the effect on inattention.

Regarding future research, our results suggest that findings from cross-sectional studies, for example those showing that subjective over objective sleep parameters predict daytime

TE D

performance, should be replicated in experience-sampling and longitudinal designs that enable an examination of current sleep. Future mediation studies should further investigate the mechanisms through which the chronotype influences functioning. According to our data, the

EP

effect does not seem to be explained by any of the state variables assessed. We can only tentatively draw conclusions regarding interventions because we did not

AC C

investigate a clinical sample (eg, adolescents with sleep or attention disorders). Nevertheless, the question arises whether daytime inattention in adolescents can be ameliorated by improving their sleep. Our results indicate that extending sleep time, for example, through better sleep hygiene [43], might have the potential to reduce inattention. However, the effect of sleep time on inattention was small. Thus, extending sleep time is unlikely to be sufficient on its own. Given that we found chronotype to predict inattention regardless of current sleep and chronotype is

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

13

generally recognized as not modifiable, it might be necessary to consider environmental changes to meet the needs of ‘owls’. As suggested by many scholars (eg, [44,45]), starting school at a later time might be recommendable for adolescents.

RI PT

Conflict of interest

We declare that no potential conflict of interest is associated with the manuscript. There

SC

was no external funding of the study.

Acknowledgements

M AN U

We would like to thank the participating adolescents, the psychology students who helped with the data collection, and Dr Jürgen Kempkensteffen for technical support.

References

4. 5. 6.

7. 8.

TE D

3.

EP

2.

Gradisar M, Gardner G, Dohnt H. Recent worldwide sleep patterns and problems during adolescence: A review and meta-analysis of age, region, and sleep. Sleep Med. 2011;12(2):110-8. National Sleep Foundation. Sleep in America® poll: Sleep in the modern family. 2014; https://sleepfoundation.org/sites/default/files/2014-NSF-Sleep-in-America-pollsummary-of-findings---FINAL-Updated-3-26-14-.pdf. Accessed 03 February, 2017. Crowley SJ, Acebo C, Carskadon MA. Sleep, circadian rhythms, and delayed phase in adolescence. Sleep Med. 2007;8(6):602-12. Bartel KA, Gradisar M, Williamson P. Protective and risk factors for adolescent sleep: A meta-analytic review. Sleep Med. Rev. 2015;21:72-85. Carskadon MA. Sleep in adolescents: The perfect storm. Pediatr Clin North Am 2011;58(3):637-47. Dewald JF, Meijer AM, Oort FJ, Kerkhof GA, Bogels SM. The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A metaanalytic review. Sleep Med. Rev. 2010;14(3):179-89. Diaz-Morales JF, Escribano C. Social jetlag, academic achievement and cognitive performance: Understanding gender/sex differences. Chronobiol. Int. 2015;32(6):822-31. Kolomeichuk SN, Randler C, Shabalina I, Fradkova L, Borisenkov M. The influence of chronotype on the academic achievement of children and adolescents – evidence from Russian Karelia. Biol. Rhythm Res. 2016;47(6):873-83.

AC C

1.

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

15.

16.

17. 18. 19. 20.

21. 22.

23.

24.

25. 26. 27.

RI PT

14.

SC

13.

M AN U

12.

TE D

11.

EP

10.

Merikanto I, Lahti T, Puusniekka R, Partonen T. Late bedtimes weaken school performance and predispose adolescents to health hazards. Sleep Med. 2013;14(11):110511. Beebe DW. Cognitive, behavioral, and functional consequences of inadequate sleep in children and adolescents. Pediatr Clin North Am. 2011;58(3):649-65. Gradisar M, Terrill G, Johnston A, Douglas P. Adolescent sleep and working memory performance. Sleep Biol. Rhythms 2008;6(3):146-54. Kopasz M, Loessl B, Hornyak M, et al. Sleep and memory in healthy children and adolescents – A critical review. Sleep Med. Rev. 2010;14(3):167-77. McGlinchey EL, Harvey AG. Risk behaviors and negative health outcomes for adolescents with late bedtimes. J Youth Adolesc. 2015;44(2):478-88. Barnes JC, Meldrum RC. The impact of sleep duration on adolescent development: A genetically informed analysis of identical twin pairs. J Youth Adolesc 2015;44(2):489506. Boschloo A, Krabbendam L, Dekker S, Lee N, de Groot R, Jolles J. Subjective sleepiness and sleep quality in adolescents are related to objective and subjective measures of school performance. Front.Psychol. 2013;4. van der Heijden KB, de Sonneville LMJ, Swaab H. Association of eveningness with problem behavior in children: A mediating role of impaired sleep. Chronobiol. Int. 2013;30(7):919-29. Adan A, Archer SN, Hidalgo MP, Di Milia L, Natale V, Randler C. Circadian typology: A comprehensive review. Chronobiol Int. 2012;29(9):1153-75. Fabbian F, Zucchi B, De Giorgi A, et al. Chronotype, gender and general health. Chronobiol Int. 2016;33(7):863-82. Randler C, Frech D. Young people's time-of-day preferences affect their school performance. J. Youth Studies 2009;12(6):653-67. Schlarb AA, Sopp R, Ambiel D, Grunwald J. Chronotype-related differences in childhood and adolescent aggression and antisocial behavior – A review of the literature. Chronobiol Int. 2014;31(1):1-16. Lange L, Randler C. Morningness-eveningness and behavioural problems in adolescents. Sleep Biol Rhythms 2011;9(1):12-8. Martin JS, Gaudreault MM, Perron M, Laberge L. Chronotype, light exposure, sleep, and daytime functioning in high school students attending morning or afternoon school shifts: An actigraphic study. J. Biol. Rhythms 2016;31(2):205-17. Hur YM. Stability of genetic influence on morningness-eveningness: A cross-sectional examination of South Korean twins from preadolescence to young adulthood. J Sleep Res. 2007;16(1):17-23. Escribano C, Diaz-Morales JF, Delgado P, Collado MJ. Morningness/eveningness and school performance among Spanish adolescents: Further evidence. Learn. Individ. Differ. 2012;22(3):409-13. Trull TJ, Ebner-Priemer U. Ambulatory assessment. Annu. Rev. Clin. Psychol. 2013;9:151-76. Meltzer LJ, Montgomery-Downs HE, Insana SP, Walsh CM. Use of actigraphy for assessment in pediatric sleep research. Sleep Med Rev. 2012;16(5):463-75. Sadeh A. The role and validity of actigraphy in sleep medicine: An update. Sleep Med. Rev. 2011;15(4):259-67.

AC C

9.

14

ACCEPTED MANUSCRIPT CHRONOTYPE, SLEEP, AND INATTENTION

34. 35. 36. 37. 38.

39.

40. 41. 42.

43. 44. 45.

RI PT

33.

SC

32.

M AN U

31.

TE D

30.

EP

29.

Randler C. German version of the reduced Morningness-Eveningness Questionnaire (rMEQ). Biol. Rhythm Res. 2013;44(5):730-6. Urban R, Magyarodi T, Rigo A. Morningness-eveningness, chronotypes and healthimpairing behaviors in adolescents. Chronobiol Int. 2011;28(3):238-47. Randler C, Freyth-Weber K, Rahafar A, Florez Jurado A, Kriegs JO. Morningnesseveningness in a large sample of German adolescents and adults. Heliyon. 2016;2(e00200). Tonetti L, Adan A, Di Milia L, Randler C, Natale V. Measures of circadian preference in childhood and adolescence: A review. Eur Psychiatry 2015;30(5):576-82. Hoffmann RM, Müller T, Hajak G, Cassel W. Abend-Morgenprotokolle in Schlafforschung und Schlafmedizin: Ein Standardinstrument für den deutschsprachigen Raum (Sleep logs in sleep research and sleep medicine). Somnologie 1997;1(3):103-9. Wilhelm P, Schoebi D. Assessing mood in daily life: Structural validity, sensitivity to change, and reliability of a short-scale to measure three basic dimensions of mood. Eur. J. Psychol. Assess. 2007;23(4):258-67. Ancoli-Israel S, Martin JL, Blackwell T, et al. The SBSM guide to actigraphy monitoring: Clinical and research applications. Behav Sleep Med. 2015;13:S4-S38. Meltzer LJ, Walsh CM, Traylor J, Westin AML. Direct comparison of two new actigraphs and polysomnography in children and adolescents. Sleep. 2012;35(1):159-66. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders – DSM-5. Washington, DC: 2013. Enders CK, Tofighi D. Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychol. Meth 2007;12(2):121-38. Baron RM, Kenny DA. The moderator-mediator varible distinction in social psychological research: Conceptual, strategic, and statistical considerations. J. Personal. Soc. Psychol. 1986;51(6):1173-82. Luciana M, Collins PF, Olson EA, Schissel AM. Tower of London performance in healthy adolescents: The development of planning skills and associations with selfreported inattention and impulsivity. Dev. Neuropsychol. 2009;34(4):461-75. Connors LL, Connolly J, Toplak ME. Self-reported inattention in early adolescence in a community sample. J. Attention Disord. 2012;16(1):60-70. Roenneberg T, Kuehnle T, Pramstaller PP, et al. A marker for the end of adolescence. Curr. Biol. 2004;14(24):R1038-9. Vollmer C, Jankowski KS, Diaz-Morales JF, Itzek-Greulich H, Wust-Ackermann P, Randler C. Morningnesse-eveningness correlates with sleep time, quality, and hygiene in secondary school students: A multilevel analysis. Sleep Med. 2017;30:151-9. Malone SK. Early to bed, early to rise? An exploration of adolescent sleep hygiene practices. J School Nurs. 2011;27(5):348-54. Owens JA, Allison M, Ancona R, et al. School start times for adolescents. Pediatrics 2014;134(3):642-9. Borlase BJ, Gander PH, Gibson RH. Effects of school start times and technology use on teenagers’ sleep: 1999–2008. Sleep Biol. Rhythms 2013;11(1):46-54.

AC C

28.

15

ACCEPTED MANUSCRIPT

Table 1. Distribution of sleep parameters, inattention, and assessment times

0

Observed range 5–15

Feeling rested

414

3

Positive affect

414

Total sleep time (h)

Mean

SD

RI PT

Missing data (%)

Morningness

Number of assessments 61

Variable

2.21

1–5

3.35

0.96

3

0–100

38.90

414

3

2.83–12.24

Sleep efficiency (%)

414

3

57–99

Wake after sleep onset (min)

414

3

1–113

39.21

17.81

Inattention

802

6

0–99

24.63

20.28

Upon awakening/before school

414

3

05:03–14:36

08:15

1:42

At noon/after school

402

6

11:37–18:26

14:15

1:30

In the evening

400

6

18:02–01:36

21:57

1:34

23.78

7.02

1.44

84.08

7.05

M AN U

AC C

EP

TE D

Assessment times

SD, standard deviation.

SC

10.48

ACCEPTED MANUSCRIPT

Within subjects

Between subjects

(1)

(2)

(3)

(4)

(5)

(6)

(1)

1. Chronotype/morningnessa

—

—

—

—

—

—

—

b

—

—

—

—

0.49**

—

—

—

0.34**

0.24**

—

—

0.08*

0.06

0.32**

—

0.13**

0.02

0.50** -0.14**

-0.10*

-0.11*

b

4. Total sleep time 5. Sleep efficiency b

6. Wake after sleep onset b

-0.15**

-0.04

EP

7. Inattention

(3)

(4)

(5)

(6)

—

—

—

—

—

—

0.42**/0.41**

—

—

—

—

—

—

0.20/0.36**

0.50**

—

—

—

—

—

0.19/0.14

0.16

-0.02

—

—

—

—

0.07/0.04

-0.18

-0.16

0.45**

—

—

—

0.06/0.01

0.12

0.02

-0.10

-0.62**

—

-0.08*

-0.26*/-0.11

-0.02

0.07

0.07

0.09

TE D

b

M AN U

b

3. Positive affect

(2)

SC

Variable

2. Feeling rested

RI PT

Table 2. Correlations of sleep parameters and inattention.

Chronotype is a dichotomous variable (evening vs neither type), morningness is the corresponding continuous variable.

b

On the within-subjects level, chronotype is a constant.

*p<0.05; **p<0.01.

AC C

a

-0.10

ACCEPTED MANUSCRIPT

b

SE

p

95 % CI

Feeling rested

-0.68

0.91

.46

-2.47, 1.00

Positive affect

-0.06

0.04

.15

-0.14, 0.02

Total sleep time

-1.52

0.67

.02

-2.83, -0.16

Wake after sleep onset

-0.02

0.06

.75

-0.13, 0.09

Day typea

-3.27

1.39

.02

Time of dayb

-0.57

1.22

.64

Feeling rested

-0.55

0.91

.55

Positive affect

-0.05

0.04

.22

Total sleep time

-1.34

0.69

.049

-2.69, 0.03

Wake after sleep onset

-0.01

0.06

.87

-0.12, 0.10

Predictor

EP

CI, confidence interval; SE, standard error. Weekday vs weekend; b morning vs afternoon inattention.

AC C

a

-5.98, -0.58 -2.3, 1.18

-2.34, 1.21

-0.14, 0.03

TE D

Controlled model

M AN U

Unrestricted model

SC

Model

RI PT

Table 3. Prediction of inattention by state sleep parameters.

Study day

0

8

Upon awakening/before school

Subjective sleep (sleep diary) • Feeling rested • Positive Affect

Return of the instruments

At noon/after school

Inattention during the morning

M AN U

Time of day the assessment took place

Baseline assessment Delivery of the instruments

Inattention during the afternoon

At night

Objective sleep (actigraphy) • Total sleep time • Wake after sleep onset

Objective sleep (actigraphy) • Total sleep time • Wake after sleep onset

AC C

EP

TE D

In the evening

Figure 1. Schedule of assessments.

RI PT

1-7

SC

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Total sleep time

Chronotype (Morningness)

-.12** -.10**

-.07* (-.07*) -.06* (-.06*) Inattention

RI PT

.08** .08**

AC C

EP

TE D

M AN U

SC

Figure 2. Model of the putative mediation between chronotype (morningness) and inattention through total sleep time. Numbers indicate standardized regression weights (above uncontrolled, below controlled for day type). The numbers in brackets represent the total effect when morningness is analyzed separately from total sleep time as predictor of inattention. *p < .05, **p < .01.

ACCEPTED MANUSCRIPT Highlights Shorter sleep predicted more self-reported inattention in adolescents.

•

Moreover, evening chronotype predicted inattention, independent of sleep time.

•

Longer sleep might alleviate inattention, but not the effect of being an ‘owl’.

AC C

EP

TE D

M AN U

SC

RI PT

•