Chronobiology International, Early Online: 1–7, (2014) ! Informa Healthcare USA, Inc. ISSN: 0742-0528 print / 1525-6073 online DOI: 10.3109/07420528.2014.933841
ORIGINAL ARTICLE
The reliability and validity of the Japanese version of the Children’s ChronoType Questionnaire (CCTQ) in preschool children Kaneyoshi Ishihara1, Yuriko Doi2, and Makoto Uchiyama3 Department of Child Welfare, Notre Dame Seishin University, Okayama, Japan, 2Area on Epidemiological Research, National Institute of Public Health, Wako, Saitama, Japan, and 3Department of Psychiatry, Nihon University School of Medicine, Tokyo, Japan
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
1
We aimed to examine the reliability and validity of the Japanese version of the Children’s ChronoType Questionnaire (CCTQ) in preschool children. The CCTQ consists of 16 items on sleep–wake parameters for scheduled and free days, a 10-item of the Morningness/Eveningness Scale (CCTQ-M/E), and a single item on chronotype. Out of 502 children aged 3–6 years living in Okayama Prefecture, we evaluated 346 (188 boys and 158 girls) between May and June 2012. Their parents filled out the questionnaires two times at an interval of two weeks. Cronbach’s a of the CCTQ-M/E was 0.77. For test–retest reliability, Pearson’s correlation coefficient of the CCTQ-M/E between the two observations was 0.898 (p50.001). Kruskal–Wallis test with post-hoc tests was used to compare sleep–wake parameters measured with the CCTQ among the three groups of children, morning (M)-type, neither (N)-type and evening (E)-type, who were classified according to the CCTQ-M/E score. Sleep–wake parameters in timing were significantly different among the children with M-type, N-type and E-type (p50.001). Post-hoc pairwise comparisons revealed that sleep–wake parameters in timing were significantly delayed from the M-type to the N-type children (p50.001), from the M-type to the E-type children (p50.001), and from the N-type to the E-type children (p50.001), except that wake-up time and get-up time were not significantly different between the children with N-type and E-type on scheduled days when their start time was regularly fixed. Out of these 346 children, we evaluated 72 (35 boys and 37 girls) to see the correlations between subjectively and objectively measured sleep–wake parameters from June to October 2012. Spearman’s correlation coefficients between sleep–wake parameters measured with the CCTQ and an actigraph were 0.512–0.836 on scheduled days (p50.001) and 0.380–0.786 on free days (p50.001). Based on these findings we conclude that the Japanese version of the CCTQ is a reliable and valid measure for assessing chronotypes in preschool children. Keywords: Chronotype, Children’s ChronoType Questionnaire, CCTQ, reliability, validity
INTRODUCTION
2012). Recently, later chronotype and larger dinner were reported to be associated with poorer glycemic control in patients with type 2 diabetes independently of sleep disturbances (Reutrakul et al., 2013). Self-report questionnaires, the Morningness– ¨ stberg, Eveningness Questionnaire (MEQ; Horne & O 1976), its revised versions (e.g. the Composite Scale of Morningness [CSM; Smith et al., 1989], the Morningness–Eveningness Scale for Children [MESC; Carskadon et al., 1993]) and the Munich ChronoType Questionnaire (MCTQ; Roenneberg et al., 2003), have been widely used to assess this individual difference in adults, adolescents, and children as young as 10 years of age. Recently, Werner et al. (2009) developed the Children’s ChronoType Questionnaire (CCTQ), an adaptation of the MCTQ (Roenneberg et al., 2003) and the
Circadian rhythmic expression differs among individuals and is classified according to circadian typology, which consists of three chronotypes: morning- (M-type), neither- (N-type), and evening-type (E-type) (Adan et al., 2012). Chronotypes reflect the time of day at which individuals are ‘‘at their best’’ (Kerkhof, 1985; Werner et al., 2009) or at their personal ‘‘feeling-best ¨ stberg, 1976; Roenneberg et al., rhythm’’ (Horne & O 2003). This individual difference affects our biological and psychological functioning, not only in terms of health but also in case of disease (Adan et al., 2012). Among the symptoms that are often associated with the presence of psychiatric symptoms and chronotype are alterations in appetite, in cognition and in activity, which together can lead social damage (Adan et al.,
Submitted April 17, 2014, Returned for revision June 7, 2014, Accepted June 09, 2014
Correspondence: Yuriko Doi, National Institute of Public Health, 2-3-6 Minami, Wako, Saitama 3510197, Japan. Tel: +8148 4586111. Fax: +8148 4690213. E-mail: [email protected]
1
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
2
K. Ishihara et al.
MESC (Carskadon et al., 1993), and provided its psychometric properties as a valid and reliable measure for prepubertal children aged 4–11 years (Werner et al., 2009). We developed the Japanese version of the CCTQ, and reported that it was reliable for preschool children who lived in the Tokyo metropolitan area (Doi et al., 2013) and their chronotypes significantly differed in sleep– wake timing (Doi et al., 2014). Based on previous studies (Adan et al., 2012; Doi et al., 2014; Horne & ¨ stberg, 1976; Kerkhof, 1985; Roenneberg et al., 2003; O Werner et al., 2009), we hypothesized that sleep–wake timing is delayed in children according to later chronotype. To examine this hypothesis, it is important to objectively assess sleep–wake time. Actigraphy is an objective, non-intrusive method for estimating sleep– wake patterns using activity-based monitoring within the child’s natural environment to delineate sleep patterns and document treatment responses in normal infants and children as well as in special pediatric populations. It is used because traditional sleep monitoring by polysomnography can be difficult to perform and/or interpret in children (Ancoli-Israel et al., 2003; Meltzer et al., 2012; Morgenthaler et al., 2007). The aims of our present study were to examine the reliability and validity of the Japanese version of the CCTQ for preschool children: to replicate the results of our aforementioned previous work by using the same method in a different sample of preschool children, and to assess the validity of its sleep–wake measurements by comparing them to those derived from actigraphic recordings.
METHODS Subjects and procedures The present study consists of two parts—Part 1 and Part 2—which examine the reliability and validity of the CCTQ in preschool children. The Ethics Committee on Human Research of Notre Dame Seishin University, Japan, approved this study. The study protocol complies with the ethical standards of the Journal for the conduct of human biological rhythm research outlined in Portaluppi et al. (2010). The details of methods used in each part are as follows. Part 1 The subjects included 502 children, aged 3–6 years, who attended two kindergartens in Okayama City and three childcare centers in Kurashiki City. Between May and June 2012, we provided teachers at these institutions with legal-sized envelopes containing an invitation letter, two forms of parent-report anonymous questionnaires including the CCTQ (Doi et al., 2013; Werner et al., 2009), an envelope, and a ballpoint pen. These were then provided to all the parents by the teachers. The parents were instructed to fill out one of the forms
and seal it in the envelope immediately after receiving it, and to fill out the other form after two weeks. In total, 358 parents agreed to participate in the study and returned the completed forms in sealed envelopes to us through the teachers. A total of 346 children (188 boys and 158 girls; 4.3 ± 0.8 (mean ± standard deviation [SD]) years old) were selected as subjects to be evaluated, after excluding 12 children who missed at least one of the items on the CCTQ.
Part 2 We sent another invitation letter to those who participated in Part 1 of our study, in which we explained the purpose and procedure of Part 2 of this study, and requested their participation. After orally confirming their interest by telephone, 71 parents (with 75 children in all) agreed to participate in Part 2 of the study. We mailed each family an envelope containing consent and withdrawal forms, a stamped and addressed envelope, instructions, a parent-report questionnaire including the CCTQ, a parent-report sleep dairy, and an actigraph. The parents were instructed to put the actigraph on their child’s non-dominant wrist at home approximately one hour before bedtime and remove it one hour after get-up time, and press its event marker at the times when their child went to bed in the evening and got up out of bed in the morning, for seven consecutive nights. Only in cases in which they developed a bedtime ritual (e.g. reading a bedtime story), they were instructed to press an event marker at the time of lights-off as well. They were also instructed to keep a daily sleep diary of bedtime/time of lights-off, sleep onset time at night and get-up time in the morning, and fill out the CCTQ at a certain point of time during the corresponding period. Part 2 of the study was conducted between June and October, except for mid-August, in 2012. Seventy-two children were evaluated (35 boys and 37 girls; 4.8 [mean] ± 0.8 [SD] years of age); three were excluded because of missing data on actigraphic recordings due to participant non-adherence. Each parent provided written consent and received rewards for participation. Measures CCTQ The CCTQ (Werner et al., 2009), a parent-report questionnaire, consists of 16 items on sleep–wake parameters for scheduled days (weekdays) and free days (weekends), a 10-item M/E scale (CCTQ-M/E; range, 10–48), and a single item on chronotype (CT) (range, 1–5). The CCTQ also includes a total score of the CCTQM/E, which is calculated by adding the 10 items. The assignment of scores for Items 1, 2, 8, and 9 in the scale are reversed. It provides a range of 10 (extreme morningness) to 48 (extreme eveningness), and three different measures of children’s chronotypes as follows: M-type, N-type and E-type, which are classified using an M/E scale score of 23, 24–32, and 33, respectively. Chronobiology International
Reliability and validity of the CCTQ
3
TABLE 1. Definitions of sleep–wake parameters measured with the Children’s ChronoType Questionnaire (CCTQ) and actigraphy. Sleep-wake parameters Timing Bedtime
Actigraphy
the time when my child goes to bed (body in bed)
Sleep offset time
the time when my child is ready to fall asleep (lights turned out) the time when sleep latency is added to time of lights-off Not applicable
Wake-up time Get-up time
the time when my child wakes up the time when my child gets up
Midsleep point
the time when sleep onset is added to sleep period/2
the time when a child goes to bed, as indicated by an event marker or a sleep diary (reported time) the time when the lights are turned out, as indicated by an event marker or a sleep diary (reported time) the time for first of consecutive 5 minutes of sleep following the time of lights-off the time for last of consecutive 5 minutes of sleep prior to reported get-up time the same as above the time when a child gets up, as indicated by an event marker or a sleep diary (reported) the time halfway between sleep onset time and sleep offset time
the duration from bedtime to get-up time the duration from sleep onset to wake-up time the duration when it takes my child to fall asleep (after lights turned out)
the duration from bedtime to get-up time (reported) the duration from sleep onset time to sleep offset time the duration from the time of lights-off to sleep onset time
Time of lights-off Sleep onset time
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
CCTQ
Duration Time in bed Sleep period Sleep latency
Time in Bed
The Japanese version of the CCTQ (Doi et al., 2013) was used for the current study. Sleep latency
Actigraphy ActiWatch2 (Philips Respironics, Murraysville, PA) was used in the present study. Data were collected in oneminute intervals using threshold automatic sensitivity and scored using Actiware software version 5.70.0 (Philips Respironics, Murraysville, PA). Concurrent sleep diaries were used to confirm sleep onset and sleep offset times, and to check artifacts appearing on the actigraphic recordings. Variables The definitions of the sleep–wake parameters measured with the CCTQ and actigraphy are summarized in Table 1 and Figure 1. Statistical analyses Part 1 To assess the internal consistency of the scale, a Cronbach’s a coefficient was calculated. A value above 0.70, but not higher than 0.9, was deemed satisfactory. To assess the homogeneity of the scale, the corrected item-total correlation was used. As a usual rule of thumb, a value above 0.2 was assumed as the acceptable one for homogeneity of the scale (Nunnally, 1994; Streiner, 1995). To examine test–retest reliability of the scale, Pearson’s correlation coefficient was used (Rousson, 2002). The sleep–wake parameters calculated from the CCTQ, most of which were not normally distributed, are presented as median and interquartile range (IQR) in hours: in minutes for clock-time, in decimal hours for TIB and sleep period, and in minutes for sleep latency. The time when the subject was fully alert and sleep inertia were not used in this study because they could !
Informa Healthcare USA, Inc.
Sleep onset time
Sleep period
Midsleep point
Wake-up time/ Sleep offset time
Time of lights-off Bedtime
Get-up time
FIGURE 1. Timing and duration of sleep–wake parameters.
not be measured by actigraphy. The sleep–wake parameters used for the analyses are shown in Table 1 and Figure 1. Kruskal–Wallis test was used to examine the differences in sleep–wake parameters among the three groups of children with chronotypes: M-type, N-type and E-type. Post-hoc tests were used to determine which of these groups differed from each other.
Part 2 The sleep–wake parameters, shown in Table 1 and Figure 1, were calculated from the CCTQ, sleep diary and actigraphy. The average sleep–wake parameters were separately computed for actigraphy for five scheduled days and for two free days. Those parameters are presented in hours as the mean, SD, median, and IQR: in minutes for clock-time and in decimal hours or minutes for duration. Spearman correlation coefficients were calculated to examine the correlations between sleep–wake parameters (the subjectively measured ones obtained by the CCTQ and the objectively measured ones obtained by actigraphy).
4
K. Ishihara et al.
The analyses of Part 1 and Part 2 in the present study were conducted with two-tailed tests, and a p value 50.05 was considered statistically significant. Computations were based on the pairwise deletion of missing data. PASW statistics 19 (SPSS, Tokyo, Japan) was used to perform all statistical analyses.
Part 1 As shown in Table 2, the overall Cronbach’s a for the CCTQ-M/E was 0.77, which was higher than the Cronbach’s a value for each item of the scale. The corrected item-total correlation coefficients were 40.2. For test–retest reliability of the CCTQ-M/E, the time interval between the first (Time 1) and second (Time 2) observations (mean (SD)) was 15.4 (3.8) days. Figure 2 shows that the total scores of the CCTQ-M/E at the two different observations, Time 1 and Time 2, were highly correlated: Pearson’s correlation coefficient (r) was TABLE 2 Homogeneity and internal consistency of the Morningness/Eveningness Scale in the Children’s ChronoType Questionnaire (CCTQ-M/E) in preschool children (N ¼ 346). CCTQ-M/E Item Item Item Item Item Item Item Item Item Item All a
1 2 3 4 5 6 7 8 9 10
Mean 1.86 2.12 2.44 2.97 1.40 3.24 3.15 2.49 3.35 2.81
SD
a
0.866 0.911 0.732 0.753 0.556 0.952 1.018 1.022 1.304 0.953
ab
Corrected item-total correlation
0.726 0.730 0.742 0.741 0.756 0.744 0.746 0.720 0.758 0.729 0.767
0.524 0.490 0.411 0.422 0.286 0.386 0.380 0.548 0.361 0.491 —
SD stands for standard deviation. bCronbach’s a coefficients are represented for all items included as well as each item deleted.
FIGURE 2. Test–retest reliability of the Morningness/Eveningness Scale in the Children’s ChronoType Questionnaire (CCTQ-M/E) at the two different times, Time 1 and Time 2, over two weeks in preschool children (n ¼ 308).
Part 2 Table 4 shows the sleep–wake parameters measured using the CCTQ and actigraphy. As shown in Table 5, the sleep–wake parameters measured with the CCTQ were highly correlated to those reported or measured using actigraphy (p50.001). The correlation coefficients were slightly smaller on free days than on scheduled days. For sleep latency on scheduled days and free days, and sleep period on free days, the correlation coefficients were smaller than those of the remaining parameters.
50 45 40 CCTQ-ME score at Time 2
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
RESULTS
0.898 (p50.001); r2 was 0.806; and the regression coefficient b was 0.900. Table 3 shows the parents’ reports of their children’s sleep–wake timing and duration according to chronotype. Sleep–wake parameters in timing were significantly different among the children according to chronotype on both scheduled and free days (p50.001). The differences in wake-up time, get-up time and midsleep point were greater on free days (H(2) ¼ 72.27–98.51, p50.001) than scheduled days (H(2) ¼ 50.20–87.08, p50.001). Post-hoc pairwise comparisons revealed that most sleep–wake parameters in timing were significantly delayed from the M-type to the N-type children (p50.001), from the M-type to the E-type children (p50.001), and from the N-type to the E-type children (p50.001), but wake-up time and getup time on scheduled days were not significantly different between the children with N-type and E-type. Post-hoc pairwise comparisons showed, on scheduled days, TIB and sleep period were significantly shorter in the E-type than in the M- or N-type children (p50.001). On free days, however, those durations did not significantly differ among the children with M-type, N-type and E-type. The duration of sleep latency was significantly longer in the E-type than in the M-type children on both scheduled (p50.05) and free (p50.01) days.
y = 2.273 + 0.900x r2= 0.806
35 30 25 20 15 10 5 0 0
5
10
15
20
25
30
35
40
45
50
CCTQ-ME score at Time 1 Chronobiology International
Reliability and validity of the CCTQ
5
TABLE 3. Comparisons of sleep–wake parameters according to chronotypes, using the Children’s ChronoType Questionnaire (CCTQ), among preschool children (N ¼ 346). M-type Sleep–wake parameters
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
Scheduled days Timing Bedtime Sleep onset time Wake-up time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency Free days Timing Bedtime Sleep onset time Wake-up time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency
Median
21:00 21:05 6:30 6:39 1:53
IRQ
58 60 30 30 36
N-type N
116 21:00 115 21:36 117 7:00 116 7:00 115 2:18
10.00 1.08 115 9.58 0.98 115 10.0 10.0 116
21:00 21:00 7:00 7:00 2:05
60 69 30 59 40
Median
10.00 1.02 115 9.54 0.99 116 10.0 10.0 116
IRQ
N
30 46 49 50 31
N
H (2)
P
Post-hoc pairwise comparison
31 29 31 30 29
64.76 69.34 50.20 54.18 87.08
*** *** *** *** ***
M5N*** M5E*** N5E*** M5N*** M5E*** N5E*** M5N***, M5E*** M5N***, M5E*** M5N***, M5E***, N5E***
9.29 1.13 30 8.92 1.00 29 15.0 25.0 30
9.82 17.17 7.08
** *** *
M4E**, N4E* M4E***, N4E* M5E*
31 28 31 31 28
62.00 62.90 84.50 72.27 98.51
*** *** *** *** ***
M5N***, M5N***, M5N***, M5N***, M5N***,
10.25 1.50 31 9.58 1.52 28 15.0 24.0 29
0.87 0.61 8.78
N.S. N.S. *
M5E**
Median
198 22:00 194 22:10 196 7:00 198 7:13 192 2:42
9.83 1.20 198 9.33 1.06 192 10.0 12.5 193
117 21:30 116 22:00 117 7:30 114 7:32 116 2:40
Kruskal–Wallisa
E-type
60 50 60 70 58
198 22:00 195 22:20 198 8:00 198 8:00 195 3:03
10.17 1.17 198 9.67 1.17 195 10.0 12.5 194
IRQ
45 40 40 30 36
60 65 60 90 37
M5E***, M5E***, M5E***, M5E***, M5E***,
N5E* N5E** N5E* N5E* N5E**
Reported as the median and interquartile range (IQR) in hours: minutes for timings and decimal hours for time in bed and sleep period, and minutes for sleep latency. Morning-(M-type), neither-(N-type), and evening-(E-type) types were classified by the score of the M/E Scale in the CCTQ (CCTQ-M/E). aThe test statistics of Kruskal–Wallis test are presnted as Kruskal–Wallis H and the dgree of freedom (H (2)) and the significance level (P). N.S.: not significant *p50.05 **p50.01 ***p50.001
TABLE 4. Sleep–wake parameters measured using the Children’s ChronoType Questionnaire (CCTQ) and actigraphy in preschool children (N ¼ 72). CCTQ Sleep-wake parameters Scheduled days Timing Bedtime Sleep onset time Wake-up time/Sleep offset time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency Free days Timing Bedtime Sleep onset time Wake-up time/Sleep offset time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency
Mean
21:08 21:32 6:48 6:54 2:10 9.77 9.28 14.8
21:22 21:45 7:16 7:26 2:30 10.08 9.52 14.3
SD
42 40 29 26 28 0.69 0.67 11.6
44 40 45 45 36 0.84 0.77 11.1
Actigraphy
Median
21:00 21:30 6:53 7:00 2:08 9.83 9.33 12.5
21:30 21:45 7:15 7:30 2:23 10.04 9.44 12.5
IQR
30 49 40 41 30 0.92 0.81 15.0
30 39 60 60 49 0.99 0.96 10.0
Mean
21:09 21:30 6:41 6:49 2:10 9.65 9.18 21.1
21:21 21:42 6:57 7:09 2:25 9.80 9.26 20.8
SD
41 41 28 27 29 0.64 0.63 11.5
43 42 43 41 37 0.75 0.73 14.6
Median
21:11 21:32 6:41 6:49 2:10 9.64 9.20 19.3
21:24 21:42 6:56 7:09 2:26 9.76 9.20 17.0
IQR
55 51 34 35 30 0.95 1.01 15.1
49 51 55 61 45 1.10 1.24 16.1
Reported as the mean and standard deviation (SD), and median and interquartile range (IQR) in hours: minutes for timings, and decimal hours for time in bed and sleep period, and minutes for sleep latency. Regarding actigraphy, means (SD) and medians (IQR) represent the average sleep–wake parameters for five scheduled days and for two free days. !
Informa Healthcare USA, Inc.
6
K. Ishihara et al.
TABLE 5. Correlations between sleep–wake parameters measured with the Children’s Chronotype Questionnaire (CCTQ) and actigraphy in preschool children (N ¼ 72). Scheduled days Sleep-wake parameter
a
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
Reported Bedtime 0.774 Get-up time 0.830 Time in bed 0.740 Actigraphy Sleep onset time 0.775 Wake-up time/Sleep offset time 0.824 Midsleep point 0.836 Sleep period 0.743 Sleep latency 0.512
Free days
p Value
a
p Value
*** *** ***
0.641 0.714 0.556
*** *** ***
*** *** *** *** ***
0.606 0.707 0.786 0.490 0.380
*** *** *** *** ***
Regarding actigraphy, sleep–wake parameters were averaged for five scheduled days and for two free days. aSpearman correlation coefficient. ***p50.001.
DISCUSSION The present study demonstrates that the CCTQ-M/E is internally reliable. The Cronbach’s a was 0.77 and the corrected item-total correlation coefficients were 40.2, which met the criterion values for homogeneity (Streiner, 1995). In addition, its test–retest reliability was corroborated by the finding that the total scores of the CCTQ-M/E at the two different times were highly correlated; the correlation coefficient was 0.898. These findings are consistent with those of our previous study on a different community sample of preschool children aged 4–6 years. That study reported that the Cronbach’s a was 0.76 and the correlation coefficient of test–retest reliability was 0.853 (Doi et al., 2013, 2014). Taken together, all the findings confirm that the CCTQ-M/E is reliable. As shown in Table 3, parent-reported sleep–wake timing was significantly delayed according to later chronotype on both scheduled and free days, except for wake-up time and get-up time on scheduled days. Children with E-type had to get up earlier than their preferred wake time to be in time for kindergarten or childcare center on scheduled days when their start time was regularly fixed. These findings are consistent with our previous study (Doi et al., 2014). We found similar results in other studies. Chronotypes as measured by the CCTQ-M/E score were significantly related to sleep– wake timing in children (Werner et al., 2009). The MESC scores were significantly correlated with weekday and weekend bedtimes as well as wake-up time in adolescents (Carskadon et al., 1993). These findings support the hypothesis that sleep–wake timing is delayed in children according to later chronotype. With regard to sleep duration, we find almost the same length of nocturnal sleep among the children with M-type, N-type and E-type on free days as their innate state but it significantly shorter in the E-type children on scheduled days in our present and previous studies
(Doi et al., 2014). These findings may reflect the misalignment of biological and social time. This so-called social jet lag (Roenneberg et al., 2012; Wittmann et al., 2006) resulted in sleep restriction on scheduled days in children with later chronotypes. The degree of this misalignment may vary, depending on the cultural values, beliefs, and practice of sleep–wake behaviors as well as daytime and nighttime activities in the society in which the children live (Jenni & O’Connor, 2005; Mindell et al., 2013). Thus, basically, children’s nocturnal sleep durations may vary according to socio-cultural conditions but remain relatively constant among chronotypes if their sleep needs are satisfied. The strength of our present study was that we examined the validity of sleep–wake parameters measured with the CCTQ by comparing them to those measured using actigraphy. Ideally, we would have investigated all children participating in Part 1 of this study, but 72 children were evaluated due to practical difficulties. As shown in Table 5, the correlations were high between the two for mid-sleep point and wake-up time/sleep offset time on scheduled days and free days, and for sleep onset time and sleep period on scheduled days; moderate correlations held for sleep latency on scheduled days and sleep onset time on free days; and low correlations held for sleep period and sleep latency on free days. These findings are almost consistent with those reported from the previous studies of adolescents (Gaina et al., 2004; Wolfson et al., 2003) and infants (Sadeh, 2004). Since an actigraph differentiates between wakefulness and sleep, based on the level of body movements, false positives when the child is sleeping but shows signs of waking (e.g. turning over or changing sleep positions) or false negatives when the child is awake but appears asleep (e.g. waking while being still) may occur (Be´langer et al., 2013; De Koninck et al., 1992; Sitnick et al., 2008). In those cases, parents’ observations may reflect actual states more accurately than actigraphic recordings. To further assess children’s wakefulness and sleep, polysomnography or videosomnography is required (Ancoli-Israel et al., 2003; Sitnick et al., 2008; Tryon, 2004). As mentioned in the introduction, the concept of chronotype means differences in the time of day at which individuals are at their best. Previous studies suggest relationships between chronotypes and academic achievements in high school and university students (Be¸soluk et al., 2011; Preckel et al., 2013), health-impairing behaviors in high school students (Urba´n et al., 2011), and aggression and anti-social behaviors in children and adolescents (Schlarb et al., 2014). Future research is required to examine whether chronotypes predict individual differences in psychological functioning and behaviors in younger children. In conclusion, the present study demonstrates that the Japanese version of the CCTQ is a reliable and valid Chronobiology International
Reliability and validity of the CCTQ measure for assessing chronotypes in preschool children aged 3–6 years.
ACKNOWLEDGEMENTS We thank the parents who participated in this study as well as the principals and teachers of two kindergartens in Okayama City and three childcare centers in Kurashiki City, Okayama, Japan.
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
DECLARATION OF INTEREST This was not an industry-supported study. Dr Uchiyama reports that his department received research support from: Astellas Pharma, Eisai, Meijiseika Pharma, Mochida Pharmaceutical, MSD, Pfizer, Sanofi, Takeda Pharmaceutical, and Yoshitomiyakuhin Corporation. He has provided consulting services to Eisai, Kao, MSD, Pfizer Japan, Sanofi, Taisho Pharmaceutical and Takeda Pharmaceutical. The other authors declare no conflicts of interest. The authors alone are responsible for the contents and writing of the paper. This work was supported by JSPS KAKENHI Grant Number 23390181 and partly by a Research Grant from the Japan Society for Promoting Science and Technology Agency (22591301, 2011–2012).
REFERENCES Adan A, Archer SN, Hidalgo MP, et al. (2012). Circadian typology: A comprehensive review. Chronobiol Int. 29:1153–75. Ancoli-Israel S, Cole R, Alessi C, et al. (2003). The role of actigraphy in the study of sleep and circadian rhythms. Sleep. 26:342–92. Be´langer ME, Bernier A, Paquet J, et al. (2013). Validating actigraphy as a measure of sleep for preschool children. J Clin Sleep Med. 9:701–6. ¨ nder I, Deveci I. (2011). Morningness-Eveningness Be¸soluk S, O preferences and academic achievement of university students. Chronobiol Int. 28:118–25. Carskadon MA, Vieira C, Acebo C. (1993). Association between puberty and delayed phase preference. Sleep. 16:258–62. De Koninck J, Lorrain D, Gagnon P. (1992). Sleep positions and position shifts in five age groups: An ontogenetic picture. Sleep. 15:143–9. Doi Y, Ishihara K, Uchiyama M, Takimoto H. (2013). Development of the Japanese Version of Choldren’s Chronotype Questionnaire. Jpn J Sleep Med. 7:486–93 (in Japanese). Doi Y, Ishihara K, Uchiyama M. (2014). Sleep/wake patterns and circadian typology in preschool children based on standardized parental self-reports. Chronobiol Int. 31:328–36. Gaina A, Sekine M, Chen X, et al. (2004). Validity of child sleep diary questionnaire among junior high school children. J Epidemiol. 14:1–4. ¨ stberg O. (1976). A self-assessment questionnaire to Horne JA, O determine morningness-eveningness in human circadian rhythms. Int J Chronobiol. 4:97–110. Jenni OG, O’Connor BB. (2005). Children’s sleep: An interplay between culture and biology. Pediatrics. 115:204–16.
!
Informa Healthcare USA, Inc.
7
Kerkhof GA. (1985). Inter-individual differences in the human circadian system: A review. Biol Psychol. 20:83–112. Meltzer LJ, Montgomery-Downs HE, Insana SP, Walsh CM. (2012). Use of actigraphy for assessment in pediatric sleep research. Sleep Med Rev. 16:463–75. Mindell JA, Sadeh A, Kwon R, Goh DY. (2013). Cross-cultural differences in the sleep of preschool children. Sleep Med. 14: 1283–9. Morgenthaler T, Alessi C, Friedman L, et al. (2007). Practice parameters for the use of actigraphy in the assessment of sleep and sleep disorders: An update for 2007. Sleep. 30:519–29. Nunnally JCBI. (1994). Psychometric theory, 3rd ed. New York: McGraw-Hill Inc. Portaluppi F, Smolensky MH, Touitou Y. (2010). Ethics and methods for biological rhythm research on animals and human beings. Chronobiol Int. 27:1911–29. Preckel F, Lipnevich AA, Boehme K, et al. (2013). Morningnesseveningness and educational outcomes: The lark has an advantage over the owl at high school. Br J Educ Psychol. 83: 114–34. Reutrakul S, Hood MM, Crowley SJ, et al. (2013). Chronotype is independently associated with glycemic control in type 2 diabetes. Diabetes Care. 36:2523–9. Roenneberg T, Allebrandt KV, Merrow M, Vetter C. (2012). Social jetlag and obesity. Curr Biol. 22:939–43. Roenneberg T, Wirz-Justice A, Merrow M. (2003). Life between clocks: Daily temporal patterns of human chronotypes. J Biol Rhythms. 18:80–90. Rousson V, Gasser T, Seifert B (2002). Assessing intrarater, interrater and test–retest reliability of continuous measurements. Stat Med. 21:3431–46. Sadeh A. (2004). A brief screening questionnaire for infant sleep problems: Validation and findings for an internet sample. Pediatrics. 113:e570–7. ¨ nwald J. (2014). ChronotypeSchlarb AA, Sopp R, Ambiel D, Gru related differences in childhood and adolescent aggression and antisocial behavior – a review of the literature. Chronobiol Int. 31:1–16. Sitnick SL, Goodlin-Jones BL, Anders TF. (2008). The use of actigraphy to study sleep disorders in preschoolers: Some concerns about detection of nighttime awakenings. Sleep. 31: 395–401. Smith CS, Reilly C, Midkiff K. (1989). Evaluation of three circadian rhythm questionnaires with suggestions for an improved measure of morningness. J Appl Psychol. 74: 728–38. Streiner DLNG. (1995). Health measurement scales, a practical guide to their development and use, 2nd ed. Oxford: Oxford University Press. Tryon WW. (2004). Issues of validity in actigraphic sleep assessment. Sleep. 27:158–65. Urba´n R, Magyaro´di T, Rigo´ A. (2011). Morningness-eveningness, chronotypes and health-impairing behaviors in adolescents. Chronobiol Int. 28:238–47. Werner H, Lebourgeois MK, Geiger A, Jenni OG. (2009). Assessment of chronotype in four- to eleven-year-old children: Reliability and validity of the Children’s Chronotype Questionnaire (CCTQ). Chronobiol Int. 26:992–1014. Wittmann M, Dinich J, Merrow M, Roenneberg T. (2006). Social jetlag: Misalignment of biological and social time. Chronobiol Int. 23:497–509. Wolfson AR, Carskadon MA, Acebo C, et al. (2003). Evidence for the validity of a sleep habits survey for adolescents. Sleep. 26: 213–16.
ORIGINAL ARTICLE
The reliability and validity of the Japanese version of the Children’s ChronoType Questionnaire (CCTQ) in preschool children Kaneyoshi Ishihara1, Yuriko Doi2, and Makoto Uchiyama3 Department of Child Welfare, Notre Dame Seishin University, Okayama, Japan, 2Area on Epidemiological Research, National Institute of Public Health, Wako, Saitama, Japan, and 3Department of Psychiatry, Nihon University School of Medicine, Tokyo, Japan
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
1
We aimed to examine the reliability and validity of the Japanese version of the Children’s ChronoType Questionnaire (CCTQ) in preschool children. The CCTQ consists of 16 items on sleep–wake parameters for scheduled and free days, a 10-item of the Morningness/Eveningness Scale (CCTQ-M/E), and a single item on chronotype. Out of 502 children aged 3–6 years living in Okayama Prefecture, we evaluated 346 (188 boys and 158 girls) between May and June 2012. Their parents filled out the questionnaires two times at an interval of two weeks. Cronbach’s a of the CCTQ-M/E was 0.77. For test–retest reliability, Pearson’s correlation coefficient of the CCTQ-M/E between the two observations was 0.898 (p50.001). Kruskal–Wallis test with post-hoc tests was used to compare sleep–wake parameters measured with the CCTQ among the three groups of children, morning (M)-type, neither (N)-type and evening (E)-type, who were classified according to the CCTQ-M/E score. Sleep–wake parameters in timing were significantly different among the children with M-type, N-type and E-type (p50.001). Post-hoc pairwise comparisons revealed that sleep–wake parameters in timing were significantly delayed from the M-type to the N-type children (p50.001), from the M-type to the E-type children (p50.001), and from the N-type to the E-type children (p50.001), except that wake-up time and get-up time were not significantly different between the children with N-type and E-type on scheduled days when their start time was regularly fixed. Out of these 346 children, we evaluated 72 (35 boys and 37 girls) to see the correlations between subjectively and objectively measured sleep–wake parameters from June to October 2012. Spearman’s correlation coefficients between sleep–wake parameters measured with the CCTQ and an actigraph were 0.512–0.836 on scheduled days (p50.001) and 0.380–0.786 on free days (p50.001). Based on these findings we conclude that the Japanese version of the CCTQ is a reliable and valid measure for assessing chronotypes in preschool children. Keywords: Chronotype, Children’s ChronoType Questionnaire, CCTQ, reliability, validity
INTRODUCTION
2012). Recently, later chronotype and larger dinner were reported to be associated with poorer glycemic control in patients with type 2 diabetes independently of sleep disturbances (Reutrakul et al., 2013). Self-report questionnaires, the Morningness– ¨ stberg, Eveningness Questionnaire (MEQ; Horne & O 1976), its revised versions (e.g. the Composite Scale of Morningness [CSM; Smith et al., 1989], the Morningness–Eveningness Scale for Children [MESC; Carskadon et al., 1993]) and the Munich ChronoType Questionnaire (MCTQ; Roenneberg et al., 2003), have been widely used to assess this individual difference in adults, adolescents, and children as young as 10 years of age. Recently, Werner et al. (2009) developed the Children’s ChronoType Questionnaire (CCTQ), an adaptation of the MCTQ (Roenneberg et al., 2003) and the
Circadian rhythmic expression differs among individuals and is classified according to circadian typology, which consists of three chronotypes: morning- (M-type), neither- (N-type), and evening-type (E-type) (Adan et al., 2012). Chronotypes reflect the time of day at which individuals are ‘‘at their best’’ (Kerkhof, 1985; Werner et al., 2009) or at their personal ‘‘feeling-best ¨ stberg, 1976; Roenneberg et al., rhythm’’ (Horne & O 2003). This individual difference affects our biological and psychological functioning, not only in terms of health but also in case of disease (Adan et al., 2012). Among the symptoms that are often associated with the presence of psychiatric symptoms and chronotype are alterations in appetite, in cognition and in activity, which together can lead social damage (Adan et al.,
Submitted April 17, 2014, Returned for revision June 7, 2014, Accepted June 09, 2014
Correspondence: Yuriko Doi, National Institute of Public Health, 2-3-6 Minami, Wako, Saitama 3510197, Japan. Tel: +8148 4586111. Fax: +8148 4690213. E-mail: [email protected]
1
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
2
K. Ishihara et al.
MESC (Carskadon et al., 1993), and provided its psychometric properties as a valid and reliable measure for prepubertal children aged 4–11 years (Werner et al., 2009). We developed the Japanese version of the CCTQ, and reported that it was reliable for preschool children who lived in the Tokyo metropolitan area (Doi et al., 2013) and their chronotypes significantly differed in sleep– wake timing (Doi et al., 2014). Based on previous studies (Adan et al., 2012; Doi et al., 2014; Horne & ¨ stberg, 1976; Kerkhof, 1985; Roenneberg et al., 2003; O Werner et al., 2009), we hypothesized that sleep–wake timing is delayed in children according to later chronotype. To examine this hypothesis, it is important to objectively assess sleep–wake time. Actigraphy is an objective, non-intrusive method for estimating sleep– wake patterns using activity-based monitoring within the child’s natural environment to delineate sleep patterns and document treatment responses in normal infants and children as well as in special pediatric populations. It is used because traditional sleep monitoring by polysomnography can be difficult to perform and/or interpret in children (Ancoli-Israel et al., 2003; Meltzer et al., 2012; Morgenthaler et al., 2007). The aims of our present study were to examine the reliability and validity of the Japanese version of the CCTQ for preschool children: to replicate the results of our aforementioned previous work by using the same method in a different sample of preschool children, and to assess the validity of its sleep–wake measurements by comparing them to those derived from actigraphic recordings.
METHODS Subjects and procedures The present study consists of two parts—Part 1 and Part 2—which examine the reliability and validity of the CCTQ in preschool children. The Ethics Committee on Human Research of Notre Dame Seishin University, Japan, approved this study. The study protocol complies with the ethical standards of the Journal for the conduct of human biological rhythm research outlined in Portaluppi et al. (2010). The details of methods used in each part are as follows. Part 1 The subjects included 502 children, aged 3–6 years, who attended two kindergartens in Okayama City and three childcare centers in Kurashiki City. Between May and June 2012, we provided teachers at these institutions with legal-sized envelopes containing an invitation letter, two forms of parent-report anonymous questionnaires including the CCTQ (Doi et al., 2013; Werner et al., 2009), an envelope, and a ballpoint pen. These were then provided to all the parents by the teachers. The parents were instructed to fill out one of the forms
and seal it in the envelope immediately after receiving it, and to fill out the other form after two weeks. In total, 358 parents agreed to participate in the study and returned the completed forms in sealed envelopes to us through the teachers. A total of 346 children (188 boys and 158 girls; 4.3 ± 0.8 (mean ± standard deviation [SD]) years old) were selected as subjects to be evaluated, after excluding 12 children who missed at least one of the items on the CCTQ.
Part 2 We sent another invitation letter to those who participated in Part 1 of our study, in which we explained the purpose and procedure of Part 2 of this study, and requested their participation. After orally confirming their interest by telephone, 71 parents (with 75 children in all) agreed to participate in Part 2 of the study. We mailed each family an envelope containing consent and withdrawal forms, a stamped and addressed envelope, instructions, a parent-report questionnaire including the CCTQ, a parent-report sleep dairy, and an actigraph. The parents were instructed to put the actigraph on their child’s non-dominant wrist at home approximately one hour before bedtime and remove it one hour after get-up time, and press its event marker at the times when their child went to bed in the evening and got up out of bed in the morning, for seven consecutive nights. Only in cases in which they developed a bedtime ritual (e.g. reading a bedtime story), they were instructed to press an event marker at the time of lights-off as well. They were also instructed to keep a daily sleep diary of bedtime/time of lights-off, sleep onset time at night and get-up time in the morning, and fill out the CCTQ at a certain point of time during the corresponding period. Part 2 of the study was conducted between June and October, except for mid-August, in 2012. Seventy-two children were evaluated (35 boys and 37 girls; 4.8 [mean] ± 0.8 [SD] years of age); three were excluded because of missing data on actigraphic recordings due to participant non-adherence. Each parent provided written consent and received rewards for participation. Measures CCTQ The CCTQ (Werner et al., 2009), a parent-report questionnaire, consists of 16 items on sleep–wake parameters for scheduled days (weekdays) and free days (weekends), a 10-item M/E scale (CCTQ-M/E; range, 10–48), and a single item on chronotype (CT) (range, 1–5). The CCTQ also includes a total score of the CCTQM/E, which is calculated by adding the 10 items. The assignment of scores for Items 1, 2, 8, and 9 in the scale are reversed. It provides a range of 10 (extreme morningness) to 48 (extreme eveningness), and three different measures of children’s chronotypes as follows: M-type, N-type and E-type, which are classified using an M/E scale score of 23, 24–32, and 33, respectively. Chronobiology International
Reliability and validity of the CCTQ
3
TABLE 1. Definitions of sleep–wake parameters measured with the Children’s ChronoType Questionnaire (CCTQ) and actigraphy. Sleep-wake parameters Timing Bedtime
Actigraphy
the time when my child goes to bed (body in bed)
Sleep offset time
the time when my child is ready to fall asleep (lights turned out) the time when sleep latency is added to time of lights-off Not applicable
Wake-up time Get-up time
the time when my child wakes up the time when my child gets up
Midsleep point
the time when sleep onset is added to sleep period/2
the time when a child goes to bed, as indicated by an event marker or a sleep diary (reported time) the time when the lights are turned out, as indicated by an event marker or a sleep diary (reported time) the time for first of consecutive 5 minutes of sleep following the time of lights-off the time for last of consecutive 5 minutes of sleep prior to reported get-up time the same as above the time when a child gets up, as indicated by an event marker or a sleep diary (reported) the time halfway between sleep onset time and sleep offset time
the duration from bedtime to get-up time the duration from sleep onset to wake-up time the duration when it takes my child to fall asleep (after lights turned out)
the duration from bedtime to get-up time (reported) the duration from sleep onset time to sleep offset time the duration from the time of lights-off to sleep onset time
Time of lights-off Sleep onset time
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
CCTQ
Duration Time in bed Sleep period Sleep latency
Time in Bed
The Japanese version of the CCTQ (Doi et al., 2013) was used for the current study. Sleep latency
Actigraphy ActiWatch2 (Philips Respironics, Murraysville, PA) was used in the present study. Data were collected in oneminute intervals using threshold automatic sensitivity and scored using Actiware software version 5.70.0 (Philips Respironics, Murraysville, PA). Concurrent sleep diaries were used to confirm sleep onset and sleep offset times, and to check artifacts appearing on the actigraphic recordings. Variables The definitions of the sleep–wake parameters measured with the CCTQ and actigraphy are summarized in Table 1 and Figure 1. Statistical analyses Part 1 To assess the internal consistency of the scale, a Cronbach’s a coefficient was calculated. A value above 0.70, but not higher than 0.9, was deemed satisfactory. To assess the homogeneity of the scale, the corrected item-total correlation was used. As a usual rule of thumb, a value above 0.2 was assumed as the acceptable one for homogeneity of the scale (Nunnally, 1994; Streiner, 1995). To examine test–retest reliability of the scale, Pearson’s correlation coefficient was used (Rousson, 2002). The sleep–wake parameters calculated from the CCTQ, most of which were not normally distributed, are presented as median and interquartile range (IQR) in hours: in minutes for clock-time, in decimal hours for TIB and sleep period, and in minutes for sleep latency. The time when the subject was fully alert and sleep inertia were not used in this study because they could !
Informa Healthcare USA, Inc.
Sleep onset time
Sleep period
Midsleep point
Wake-up time/ Sleep offset time
Time of lights-off Bedtime
Get-up time
FIGURE 1. Timing and duration of sleep–wake parameters.
not be measured by actigraphy. The sleep–wake parameters used for the analyses are shown in Table 1 and Figure 1. Kruskal–Wallis test was used to examine the differences in sleep–wake parameters among the three groups of children with chronotypes: M-type, N-type and E-type. Post-hoc tests were used to determine which of these groups differed from each other.
Part 2 The sleep–wake parameters, shown in Table 1 and Figure 1, were calculated from the CCTQ, sleep diary and actigraphy. The average sleep–wake parameters were separately computed for actigraphy for five scheduled days and for two free days. Those parameters are presented in hours as the mean, SD, median, and IQR: in minutes for clock-time and in decimal hours or minutes for duration. Spearman correlation coefficients were calculated to examine the correlations between sleep–wake parameters (the subjectively measured ones obtained by the CCTQ and the objectively measured ones obtained by actigraphy).
4
K. Ishihara et al.
The analyses of Part 1 and Part 2 in the present study were conducted with two-tailed tests, and a p value 50.05 was considered statistically significant. Computations were based on the pairwise deletion of missing data. PASW statistics 19 (SPSS, Tokyo, Japan) was used to perform all statistical analyses.
Part 1 As shown in Table 2, the overall Cronbach’s a for the CCTQ-M/E was 0.77, which was higher than the Cronbach’s a value for each item of the scale. The corrected item-total correlation coefficients were 40.2. For test–retest reliability of the CCTQ-M/E, the time interval between the first (Time 1) and second (Time 2) observations (mean (SD)) was 15.4 (3.8) days. Figure 2 shows that the total scores of the CCTQ-M/E at the two different observations, Time 1 and Time 2, were highly correlated: Pearson’s correlation coefficient (r) was TABLE 2 Homogeneity and internal consistency of the Morningness/Eveningness Scale in the Children’s ChronoType Questionnaire (CCTQ-M/E) in preschool children (N ¼ 346). CCTQ-M/E Item Item Item Item Item Item Item Item Item Item All a
1 2 3 4 5 6 7 8 9 10
Mean 1.86 2.12 2.44 2.97 1.40 3.24 3.15 2.49 3.35 2.81
SD
a
0.866 0.911 0.732 0.753 0.556 0.952 1.018 1.022 1.304 0.953
ab
Corrected item-total correlation
0.726 0.730 0.742 0.741 0.756 0.744 0.746 0.720 0.758 0.729 0.767
0.524 0.490 0.411 0.422 0.286 0.386 0.380 0.548 0.361 0.491 —
SD stands for standard deviation. bCronbach’s a coefficients are represented for all items included as well as each item deleted.
FIGURE 2. Test–retest reliability of the Morningness/Eveningness Scale in the Children’s ChronoType Questionnaire (CCTQ-M/E) at the two different times, Time 1 and Time 2, over two weeks in preschool children (n ¼ 308).
Part 2 Table 4 shows the sleep–wake parameters measured using the CCTQ and actigraphy. As shown in Table 5, the sleep–wake parameters measured with the CCTQ were highly correlated to those reported or measured using actigraphy (p50.001). The correlation coefficients were slightly smaller on free days than on scheduled days. For sleep latency on scheduled days and free days, and sleep period on free days, the correlation coefficients were smaller than those of the remaining parameters.
50 45 40 CCTQ-ME score at Time 2
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
RESULTS
0.898 (p50.001); r2 was 0.806; and the regression coefficient b was 0.900. Table 3 shows the parents’ reports of their children’s sleep–wake timing and duration according to chronotype. Sleep–wake parameters in timing were significantly different among the children according to chronotype on both scheduled and free days (p50.001). The differences in wake-up time, get-up time and midsleep point were greater on free days (H(2) ¼ 72.27–98.51, p50.001) than scheduled days (H(2) ¼ 50.20–87.08, p50.001). Post-hoc pairwise comparisons revealed that most sleep–wake parameters in timing were significantly delayed from the M-type to the N-type children (p50.001), from the M-type to the E-type children (p50.001), and from the N-type to the E-type children (p50.001), but wake-up time and getup time on scheduled days were not significantly different between the children with N-type and E-type. Post-hoc pairwise comparisons showed, on scheduled days, TIB and sleep period were significantly shorter in the E-type than in the M- or N-type children (p50.001). On free days, however, those durations did not significantly differ among the children with M-type, N-type and E-type. The duration of sleep latency was significantly longer in the E-type than in the M-type children on both scheduled (p50.05) and free (p50.01) days.
y = 2.273 + 0.900x r2= 0.806
35 30 25 20 15 10 5 0 0
5
10
15
20
25
30
35
40
45
50
CCTQ-ME score at Time 1 Chronobiology International
Reliability and validity of the CCTQ
5
TABLE 3. Comparisons of sleep–wake parameters according to chronotypes, using the Children’s ChronoType Questionnaire (CCTQ), among preschool children (N ¼ 346). M-type Sleep–wake parameters
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
Scheduled days Timing Bedtime Sleep onset time Wake-up time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency Free days Timing Bedtime Sleep onset time Wake-up time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency
Median
21:00 21:05 6:30 6:39 1:53
IRQ
58 60 30 30 36
N-type N
116 21:00 115 21:36 117 7:00 116 7:00 115 2:18
10.00 1.08 115 9.58 0.98 115 10.0 10.0 116
21:00 21:00 7:00 7:00 2:05
60 69 30 59 40
Median
10.00 1.02 115 9.54 0.99 116 10.0 10.0 116
IRQ
N
30 46 49 50 31
N
H (2)
P
Post-hoc pairwise comparison
31 29 31 30 29
64.76 69.34 50.20 54.18 87.08
*** *** *** *** ***
M5N*** M5E*** N5E*** M5N*** M5E*** N5E*** M5N***, M5E*** M5N***, M5E*** M5N***, M5E***, N5E***
9.29 1.13 30 8.92 1.00 29 15.0 25.0 30
9.82 17.17 7.08
** *** *
M4E**, N4E* M4E***, N4E* M5E*
31 28 31 31 28
62.00 62.90 84.50 72.27 98.51
*** *** *** *** ***
M5N***, M5N***, M5N***, M5N***, M5N***,
10.25 1.50 31 9.58 1.52 28 15.0 24.0 29
0.87 0.61 8.78
N.S. N.S. *
M5E**
Median
198 22:00 194 22:10 196 7:00 198 7:13 192 2:42
9.83 1.20 198 9.33 1.06 192 10.0 12.5 193
117 21:30 116 22:00 117 7:30 114 7:32 116 2:40
Kruskal–Wallisa
E-type
60 50 60 70 58
198 22:00 195 22:20 198 8:00 198 8:00 195 3:03
10.17 1.17 198 9.67 1.17 195 10.0 12.5 194
IRQ
45 40 40 30 36
60 65 60 90 37
M5E***, M5E***, M5E***, M5E***, M5E***,
N5E* N5E** N5E* N5E* N5E**
Reported as the median and interquartile range (IQR) in hours: minutes for timings and decimal hours for time in bed and sleep period, and minutes for sleep latency. Morning-(M-type), neither-(N-type), and evening-(E-type) types were classified by the score of the M/E Scale in the CCTQ (CCTQ-M/E). aThe test statistics of Kruskal–Wallis test are presnted as Kruskal–Wallis H and the dgree of freedom (H (2)) and the significance level (P). N.S.: not significant *p50.05 **p50.01 ***p50.001
TABLE 4. Sleep–wake parameters measured using the Children’s ChronoType Questionnaire (CCTQ) and actigraphy in preschool children (N ¼ 72). CCTQ Sleep-wake parameters Scheduled days Timing Bedtime Sleep onset time Wake-up time/Sleep offset time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency Free days Timing Bedtime Sleep onset time Wake-up time/Sleep offset time Get-up time Midsleep point Duration Time in bed Sleep period Sleep latency
Mean
21:08 21:32 6:48 6:54 2:10 9.77 9.28 14.8
21:22 21:45 7:16 7:26 2:30 10.08 9.52 14.3
SD
42 40 29 26 28 0.69 0.67 11.6
44 40 45 45 36 0.84 0.77 11.1
Actigraphy
Median
21:00 21:30 6:53 7:00 2:08 9.83 9.33 12.5
21:30 21:45 7:15 7:30 2:23 10.04 9.44 12.5
IQR
30 49 40 41 30 0.92 0.81 15.0
30 39 60 60 49 0.99 0.96 10.0
Mean
21:09 21:30 6:41 6:49 2:10 9.65 9.18 21.1
21:21 21:42 6:57 7:09 2:25 9.80 9.26 20.8
SD
41 41 28 27 29 0.64 0.63 11.5
43 42 43 41 37 0.75 0.73 14.6
Median
21:11 21:32 6:41 6:49 2:10 9.64 9.20 19.3
21:24 21:42 6:56 7:09 2:26 9.76 9.20 17.0
IQR
55 51 34 35 30 0.95 1.01 15.1
49 51 55 61 45 1.10 1.24 16.1
Reported as the mean and standard deviation (SD), and median and interquartile range (IQR) in hours: minutes for timings, and decimal hours for time in bed and sleep period, and minutes for sleep latency. Regarding actigraphy, means (SD) and medians (IQR) represent the average sleep–wake parameters for five scheduled days and for two free days. !
Informa Healthcare USA, Inc.
6
K. Ishihara et al.
TABLE 5. Correlations between sleep–wake parameters measured with the Children’s Chronotype Questionnaire (CCTQ) and actigraphy in preschool children (N ¼ 72). Scheduled days Sleep-wake parameter
a
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
Reported Bedtime 0.774 Get-up time 0.830 Time in bed 0.740 Actigraphy Sleep onset time 0.775 Wake-up time/Sleep offset time 0.824 Midsleep point 0.836 Sleep period 0.743 Sleep latency 0.512
Free days
p Value
a
p Value
*** *** ***
0.641 0.714 0.556
*** *** ***
*** *** *** *** ***
0.606 0.707 0.786 0.490 0.380
*** *** *** *** ***
Regarding actigraphy, sleep–wake parameters were averaged for five scheduled days and for two free days. aSpearman correlation coefficient. ***p50.001.
DISCUSSION The present study demonstrates that the CCTQ-M/E is internally reliable. The Cronbach’s a was 0.77 and the corrected item-total correlation coefficients were 40.2, which met the criterion values for homogeneity (Streiner, 1995). In addition, its test–retest reliability was corroborated by the finding that the total scores of the CCTQ-M/E at the two different times were highly correlated; the correlation coefficient was 0.898. These findings are consistent with those of our previous study on a different community sample of preschool children aged 4–6 years. That study reported that the Cronbach’s a was 0.76 and the correlation coefficient of test–retest reliability was 0.853 (Doi et al., 2013, 2014). Taken together, all the findings confirm that the CCTQ-M/E is reliable. As shown in Table 3, parent-reported sleep–wake timing was significantly delayed according to later chronotype on both scheduled and free days, except for wake-up time and get-up time on scheduled days. Children with E-type had to get up earlier than their preferred wake time to be in time for kindergarten or childcare center on scheduled days when their start time was regularly fixed. These findings are consistent with our previous study (Doi et al., 2014). We found similar results in other studies. Chronotypes as measured by the CCTQ-M/E score were significantly related to sleep– wake timing in children (Werner et al., 2009). The MESC scores were significantly correlated with weekday and weekend bedtimes as well as wake-up time in adolescents (Carskadon et al., 1993). These findings support the hypothesis that sleep–wake timing is delayed in children according to later chronotype. With regard to sleep duration, we find almost the same length of nocturnal sleep among the children with M-type, N-type and E-type on free days as their innate state but it significantly shorter in the E-type children on scheduled days in our present and previous studies
(Doi et al., 2014). These findings may reflect the misalignment of biological and social time. This so-called social jet lag (Roenneberg et al., 2012; Wittmann et al., 2006) resulted in sleep restriction on scheduled days in children with later chronotypes. The degree of this misalignment may vary, depending on the cultural values, beliefs, and practice of sleep–wake behaviors as well as daytime and nighttime activities in the society in which the children live (Jenni & O’Connor, 2005; Mindell et al., 2013). Thus, basically, children’s nocturnal sleep durations may vary according to socio-cultural conditions but remain relatively constant among chronotypes if their sleep needs are satisfied. The strength of our present study was that we examined the validity of sleep–wake parameters measured with the CCTQ by comparing them to those measured using actigraphy. Ideally, we would have investigated all children participating in Part 1 of this study, but 72 children were evaluated due to practical difficulties. As shown in Table 5, the correlations were high between the two for mid-sleep point and wake-up time/sleep offset time on scheduled days and free days, and for sleep onset time and sleep period on scheduled days; moderate correlations held for sleep latency on scheduled days and sleep onset time on free days; and low correlations held for sleep period and sleep latency on free days. These findings are almost consistent with those reported from the previous studies of adolescents (Gaina et al., 2004; Wolfson et al., 2003) and infants (Sadeh, 2004). Since an actigraph differentiates between wakefulness and sleep, based on the level of body movements, false positives when the child is sleeping but shows signs of waking (e.g. turning over or changing sleep positions) or false negatives when the child is awake but appears asleep (e.g. waking while being still) may occur (Be´langer et al., 2013; De Koninck et al., 1992; Sitnick et al., 2008). In those cases, parents’ observations may reflect actual states more accurately than actigraphic recordings. To further assess children’s wakefulness and sleep, polysomnography or videosomnography is required (Ancoli-Israel et al., 2003; Sitnick et al., 2008; Tryon, 2004). As mentioned in the introduction, the concept of chronotype means differences in the time of day at which individuals are at their best. Previous studies suggest relationships between chronotypes and academic achievements in high school and university students (Be¸soluk et al., 2011; Preckel et al., 2013), health-impairing behaviors in high school students (Urba´n et al., 2011), and aggression and anti-social behaviors in children and adolescents (Schlarb et al., 2014). Future research is required to examine whether chronotypes predict individual differences in psychological functioning and behaviors in younger children. In conclusion, the present study demonstrates that the Japanese version of the CCTQ is a reliable and valid Chronobiology International
Reliability and validity of the CCTQ measure for assessing chronotypes in preschool children aged 3–6 years.
ACKNOWLEDGEMENTS We thank the parents who participated in this study as well as the principals and teachers of two kindergartens in Okayama City and three childcare centers in Kurashiki City, Okayama, Japan.
Chronobiol Int Downloaded from informahealthcare.com by University of Laval on 07/13/14 For personal use only.
DECLARATION OF INTEREST This was not an industry-supported study. Dr Uchiyama reports that his department received research support from: Astellas Pharma, Eisai, Meijiseika Pharma, Mochida Pharmaceutical, MSD, Pfizer, Sanofi, Takeda Pharmaceutical, and Yoshitomiyakuhin Corporation. He has provided consulting services to Eisai, Kao, MSD, Pfizer Japan, Sanofi, Taisho Pharmaceutical and Takeda Pharmaceutical. The other authors declare no conflicts of interest. The authors alone are responsible for the contents and writing of the paper. This work was supported by JSPS KAKENHI Grant Number 23390181 and partly by a Research Grant from the Japan Society for Promoting Science and Technology Agency (22591301, 2011–2012).
REFERENCES Adan A, Archer SN, Hidalgo MP, et al. (2012). Circadian typology: A comprehensive review. Chronobiol Int. 29:1153–75. Ancoli-Israel S, Cole R, Alessi C, et al. (2003). The role of actigraphy in the study of sleep and circadian rhythms. Sleep. 26:342–92. Be´langer ME, Bernier A, Paquet J, et al. (2013). Validating actigraphy as a measure of sleep for preschool children. J Clin Sleep Med. 9:701–6. ¨ nder I, Deveci I. (2011). Morningness-Eveningness Be¸soluk S, O preferences and academic achievement of university students. Chronobiol Int. 28:118–25. Carskadon MA, Vieira C, Acebo C. (1993). Association between puberty and delayed phase preference. Sleep. 16:258–62. De Koninck J, Lorrain D, Gagnon P. (1992). Sleep positions and position shifts in five age groups: An ontogenetic picture. Sleep. 15:143–9. Doi Y, Ishihara K, Uchiyama M, Takimoto H. (2013). Development of the Japanese Version of Choldren’s Chronotype Questionnaire. Jpn J Sleep Med. 7:486–93 (in Japanese). Doi Y, Ishihara K, Uchiyama M. (2014). Sleep/wake patterns and circadian typology in preschool children based on standardized parental self-reports. Chronobiol Int. 31:328–36. Gaina A, Sekine M, Chen X, et al. (2004). Validity of child sleep diary questionnaire among junior high school children. J Epidemiol. 14:1–4. ¨ stberg O. (1976). A self-assessment questionnaire to Horne JA, O determine morningness-eveningness in human circadian rhythms. Int J Chronobiol. 4:97–110. Jenni OG, O’Connor BB. (2005). Children’s sleep: An interplay between culture and biology. Pediatrics. 115:204–16.
!
Informa Healthcare USA, Inc.
7
Kerkhof GA. (1985). Inter-individual differences in the human circadian system: A review. Biol Psychol. 20:83–112. Meltzer LJ, Montgomery-Downs HE, Insana SP, Walsh CM. (2012). Use of actigraphy for assessment in pediatric sleep research. Sleep Med Rev. 16:463–75. Mindell JA, Sadeh A, Kwon R, Goh DY. (2013). Cross-cultural differences in the sleep of preschool children. Sleep Med. 14: 1283–9. Morgenthaler T, Alessi C, Friedman L, et al. (2007). Practice parameters for the use of actigraphy in the assessment of sleep and sleep disorders: An update for 2007. Sleep. 30:519–29. Nunnally JCBI. (1994). Psychometric theory, 3rd ed. New York: McGraw-Hill Inc. Portaluppi F, Smolensky MH, Touitou Y. (2010). Ethics and methods for biological rhythm research on animals and human beings. Chronobiol Int. 27:1911–29. Preckel F, Lipnevich AA, Boehme K, et al. (2013). Morningnesseveningness and educational outcomes: The lark has an advantage over the owl at high school. Br J Educ Psychol. 83: 114–34. Reutrakul S, Hood MM, Crowley SJ, et al. (2013). Chronotype is independently associated with glycemic control in type 2 diabetes. Diabetes Care. 36:2523–9. Roenneberg T, Allebrandt KV, Merrow M, Vetter C. (2012). Social jetlag and obesity. Curr Biol. 22:939–43. Roenneberg T, Wirz-Justice A, Merrow M. (2003). Life between clocks: Daily temporal patterns of human chronotypes. J Biol Rhythms. 18:80–90. Rousson V, Gasser T, Seifert B (2002). Assessing intrarater, interrater and test–retest reliability of continuous measurements. Stat Med. 21:3431–46. Sadeh A. (2004). A brief screening questionnaire for infant sleep problems: Validation and findings for an internet sample. Pediatrics. 113:e570–7. ¨ nwald J. (2014). ChronotypeSchlarb AA, Sopp R, Ambiel D, Gru related differences in childhood and adolescent aggression and antisocial behavior – a review of the literature. Chronobiol Int. 31:1–16. Sitnick SL, Goodlin-Jones BL, Anders TF. (2008). The use of actigraphy to study sleep disorders in preschoolers: Some concerns about detection of nighttime awakenings. Sleep. 31: 395–401. Smith CS, Reilly C, Midkiff K. (1989). Evaluation of three circadian rhythm questionnaires with suggestions for an improved measure of morningness. J Appl Psychol. 74: 728–38. Streiner DLNG. (1995). Health measurement scales, a practical guide to their development and use, 2nd ed. Oxford: Oxford University Press. Tryon WW. (2004). Issues of validity in actigraphic sleep assessment. Sleep. 27:158–65. Urba´n R, Magyaro´di T, Rigo´ A. (2011). Morningness-eveningness, chronotypes and health-impairing behaviors in adolescents. Chronobiol Int. 28:238–47. Werner H, Lebourgeois MK, Geiger A, Jenni OG. (2009). Assessment of chronotype in four- to eleven-year-old children: Reliability and validity of the Children’s Chronotype Questionnaire (CCTQ). Chronobiol Int. 26:992–1014. Wittmann M, Dinich J, Merrow M, Roenneberg T. (2006). Social jetlag: Misalignment of biological and social time. Chronobiol Int. 23:497–509. Wolfson AR, Carskadon MA, Acebo C, et al. (2003). Evidence for the validity of a sleep habits survey for adolescents. Sleep. 26: 213–16.
