Sleep Quality And Chronotype Of Portuguese School-Aged Children

The perceptions of children’s sleep quality and effective sleep quality

The first part of this paper comprises a small part of a cross-sectional study on the sleep quality of Portuguese children, and examines the sleep quality perception that school-aged children have. The Pittsburgh Sleep Quality Index (PSQI) was administered to a total of 1109 Portuguese children aged 6 to 11 years. Subsequently, the parents' evaluation of the quality of their children's sleep was analysed. To this end, the Children’s Sleep Questionnaire (CSHQ-PT) was used and a total of 883 parents answered. Our study also analysed different variables such as differences between state schools (n= 403) and private schools (n= 480) and between boys (n= 437) and girls (n= 446).

Our study is based on the concept of sleep quality. However, we have to differentiate it from sleep hygiene, since the latter includes a variety of different practices and habits that are necessary in order to have good quality of nocturnal sleep and allow daytime alertness. Sleep quality and sleep duration are two different domains of sleep, although both overlap and differentiate themselves. Sleep quality refers to the subjective indexes of how sleep is experienced, including the feeling of satisfaction and feeling rested upon awakening. The quality of sleep influences more the possibility of daytime drowsiness, changes in the emotional and behavioural states and in the cognitive functions (Dewald, Meijer, Oort, Kerkhof, & Bogels, 2010).

The adequate sleep duration for healthy individuals and for those who do not suffer from any sleep disorder varies greatly in terms of the developmental cycle; for new-borns, the adequate sleep duration is between 14 and 17 hours, for small children it is between 12 and 15 hours, for pre-school age children it is between 10 and 13 hours and for school-age children it is between 9 and 11 hours. For adolescents, the adequate sleep duration ranges from 8 to 10 hours and this is considered appropriate; for adults and young adults it is between 7 and 9 hours and for older adults it is 7 to 8 hours (Hirshkowitz, Whiton, Steven, Alessi, Oliviero, DonCarlos, Hazen, Herman, & Katz, 2015). Thus, sufficient sleep duration requirements vary throughout the life span and from person to person. Individuals who usually sleep out of the normal pattern may be exhibiting signs or symptoms of serious health problems and this may compromise their well-being (Hirshkowitz et al. 2015).

Children generally have a better quality of sleep when there are domestic rules and regular sleep-wake routines. The adequate sleep quantity and quality will only exist if well-established sleep hygiene rules (no caffeine consumption and regular bedtime hours) are applied (Buxton et al., 2015). On the other hand, poor sleep quality is more likely to exist when parents and children assume the presence of electronic gadgets in the bedroom. Therefore, it is very important to concentrate efforts on reducing the invasion of technology in the children's bedroom (Brockmanna, Diaza, Damianic, Villarroeld, Núñezc, & Brunie, 2016).

Good sleep hygiene practices are associated with better sleep in various age groups. Some authors recommend that children of all ages should go to bed before 9:00 pm, have an established bedtime routine, including reading, sleep in rooms without televisions and abstain from caffeine intake (Mindella, Meltzerb, Carskadonc, & Chervind, 2009).

In the research conducted by Silva, Barbosa, Silva and Neto (2014), the Children's Sleep Habits Questionnaire (CSHQ-PT) was adapted and validated for Portuguese culture, and a broad characterization of sleep habits in children aged 2 to 10 years was made. The results show that during the week children go to bed on average at 9:44 pm. The total daily sleep time decreases with age, and on weekends children sleep an extra hour. The need to fall asleep or sleep with the parents decreases with age. The parents' sleep quality assessment indicates that children have a prevalence of sleep problems of 10.4%, with no significant differences between age groups, parental educational subgroups, nor between mid-high and low population density (Silva, Barbosa, Silva, & Neto, 2014). The sleep problems identified by these authors refer to insufficient sleep duration, resistance to going to bed, difficulty in falling asleep alone, frequent nocturnal awakenings, and occurrence of parasomnias. The Sleep Disturbance Index (SDI) was higher in children from families with lower educational levels. This study also revealed the relationship between school performance and sleep problems. The existence of a high incidence of daytime drowsiness and sleep deprivation is also prominent in these children (Silva, Barbosa, Silva, & Neto, 2014).

The research carried out by Arriaga, Brito, Gaspar and Luz (2015) with the Children's Sleep Habits Questionnaire (CSHQ-PT) showed results with a SDI mean well above the cut-off point and the average of other studies that also used the CSHQ. It is noteworthy that both boys and girls presented a SDI above 44, with a higher mean value in females. This study also reports that there is a disparity between the effective sleep quality in children and the perception of the sleep quality that parents have of their children. These authors conclude that the studied population has a high incidence of sleep disorders. In the evaluation of the different items there are also age-related differences, showing that the disorders vary according to the stages of development (Arriaga, Brito, Gaspar, & Luz, 2015). However, more robust research is needed to better understand the causal contributions of healthy sleep practices to the onset and maintenance of children's sleep problems (Allena, Howletta, Coulombeb, & Corkuma, 2016).

The perceptions of children’s chronotype and effective cognitive preferences

The second part of this paper presents an excerpt from other cross-sectional study with Portuguese emigrant population in schools in Luxembourg with the aim of first evaluating, using one of the chronotype scales of the Children's Chronotype Questionnaire (CCTQ), the perception that Portuguese parents (residing in Luxembourg) have as to the chronotype of their children; second, it compares the perceived perception with the real chronotype of children descending from Portuguese families and inserted in school contexts with differentiated schedules. Regarding the concept of chronotype, in the areas of Psychology and Chronobiology, there is some disagreement about its designation because of the near terminology - diurnal type.

However, in our study we use the term chronotype and first we definite it. Chronotype refers to the preference of individuals regarding certain periods of the day to perform specific activities, and this preference is biologically determined and not related to social impositions or conveniences. Thus, during the day it is common to note a fluctuation of cognitive performance, which influences the performance and mood of individuals (Rocha, 2017; Schmidt, Collette, Cajochen, & Peigneux, 2007; Yoon, May & Hasher, 2000). According to how we biologically react to the organization of time, we cannot separate it from health, either biologically or psychologically, so we find time in the development and maturity of biological functions, in their internal and external expression. Understanding these aspects that are present in the occurrences lived during the 24 hours of the day is essential to be able to analyse each individual and identify his/her chronotype examined through the circadian rhythm (Rocha, 2017).

Each circadian rhythm has a profile that is genetically shaped and there are some genes responsible for the different sleep-wake rhythms, so there are people who sleep more and others who sleep less, some who prefer to go to sleep later while others prefer to go to sleep early, differentiating preferences from sleeping and waking hours (Almeida, 2013; Randler, 2009; Menna-Barreto, 2003). The term "circadian rhythmicity" differs among individuals, and it is a personality trait that shows interdependent biological correlations and appears to correlate with various types of performance. Individual differences in circadian rhythms, commonly referred to as morningness and eveningness types, indicate preferences associated with morning or night activities. The literature identifies three chronographic types or chronotype: morning, intermediate or evening (Adan, Archer, Hidalgo et al., 2012). Morning individuals prefer to go to bed and get up early, peaking before noon, and evening people, in turn, prefer to go to bed and wake up later, and perform better at the end of the day (Hines, Feng, Emmett et al., 2013; Adan et al., 2012). Since the chronotype differs in sleep-wake patterns, we consider that it is extremely important to verify the perception of parents and schools about the children's chronotype to understand academic underachievement at specific times of the school day. It is also important that this study contributes to scientific research on the performance rhythms and their educational and psychological implications, specifically in relation to school-aged populations of Portuguese ancestry or native Portuguese.

The results will be discussed based on the recent literature and reflect on the educational and psychological implications: are the school schedules adjusted according to the migrant school minorities, in this case the Portuguese minorities, in foreign schools? Moreover, the data points to a worrying lack of knowledge about the children's diurnal preferences and the inferred cognitive implications, which shows a direct correlation with aspects of their academic performance and their adaptation to the social habits of the countries where they live.

Participants

Participants are divided into two major groups; one of these groups consists of the children who responded to the Pittsburgh Sleep Quality Index (PSQI) (n= 1109), divided into 546 girls and 563 boys from state schools (n=538) and private schools (n=571). The other group comprises the parents who responded to the Children’s Sleep Habits Questionnaire (CSHQ-PT), totalling 883, given that not all parents of the 1109 children returned the questionnaires. These 883 children assessed by their parents for sleep quality are divided into 437 (49.5%) boys and 446 girls (50.5%). The mean age is 8 years ± 1.241 and 99.5% of the participants are between 6 and 10 years of age. Table 1 shows the descriptive statistics for children: gender, school and year of schooling.

Instrument and Procedures

The Pittsburgh Sleep Quality Index (PSQI) developed by Buysse, Reynolds, Monk, Berman and Kupfer (1989) and adapted in the Portuguese version by Ramalho (2008) was used to study the quality of sleep evaluated by the children. Its translation into Portuguese was done by Bertolazi also in 2008. It is considered to be stable to evaluate sleep quality, since it is accessible and identifies whether the studied individuals sleep well or poorly (Córdoba & Schmalbach, 2005). Structurally, it consists of 19 questions, with 15 multiple choice items that refer to the frequency of sleep disturbances and subjective quality of sleep, and 4 items that refer to bedtime and wake up hours and sleep latency and duration. It is considered an easy-to-use reliable instrument for clinical research in different age groups (between 6 and 90 years of age), including clinical and non-clinical populations (Duarte, 2007).

The instrument used to study the children’s effective sleep quality was the Children’s Sleep Habits Questionnaire (CSHQ-PT), which was filled by their parents. A total of 883 parents returned the questionnaires. Silva, Barbosa, Silva and Neto (2014) adapted and validated the Children's Sleep Habits Questionnaire (CSHQ) for the Portuguese population, making a broad characterization of sleep habits in children from 2 to 10 years of age. The CSHQ-PT internal consistency (Cronbach's α) was 0.78 for the total scale and ranged from 0.44 to 0.74 for subscales. The reliability test-retest for subscales (Pearson's correlations, n= 58) ranged from 0.59 to 0.85. The CSHQ-PT showed psychometric properties that are comparable to the versions of other countries and suitable for the screening of sleep disorders in children (Silva, Barbosa, Silva, & Neto, 2014).

The processing and analysis of the statistical data was made using the SPSS programme, version 23. Descriptive statistics with mean, standard deviation, asymmetry and kurtosis of the 7 components and overall PSQI in the global sample were made. For the comparison between the two groups, the non-parametric Mann-Whitney U test was performed. The Kruskal-Wallis non-parametric test was used to compare more groups; 3 degrees of freedom. In order to analyse the Children's Sleep Habits Questionnaire (CSHQ-PT), descriptive statistics were also used, with mean, standard deviation, asymmetry and kurtosis of the 33 CSHQ-PT items in the global sample. To compare the two groups, the non-parametric Mann-Whitney U test was used. The Kruskal-Wallis non-parametric H test was used again to compare more groups; 3 degrees of freedom. However, it should be noted that in this paper we only present a small part of the data analysis conducted.

Participants

Three groups of participants: Portuguese, Luxembourgish and "mixed". The group of Portuguese whose parents (mother and father) were born in Portugal, the dual nationality group whose parents are Portuguese and were born in Luxembourg and, finally, the group of Luxembourgers whose parents are Luxembourgers and born in Luxembourg. For this specific study only the first two groups were considered. Regarding the place of birth of the mothers, 71 (41.7%) were born in Luxembourg, 55 (32.3%) were born in Portugal and 44 (26%) were born in another country. As for fathers, 73 (43%) were born in Luxembourg, 58 (34%) in Portugal and 39 (23%) in another country. All children (3-12 years old) attend two schools in Northern Luxembourg. Table 2 shows the descriptive statistics of age and year of schooling for the children.

* The education system used was the education system in Luxembourg. Compulsory schooling consists of 2 years of pre-school education, followed by 6 years of primary education, equivalent to the 1st and 2nd cycles of basic education in Portugal and 3 years of secondary education (years 7, 8 and 9). Basic education (primary education) consists of 4 cycles. The 1st cycle corresponds to preschool from 4 to 5 years of age; the 2nd cycle corresponds to years 1 and 2 from 6 to 7 years of age; the 3rd cycle corresponds to years 3 and 4 from 8 to 9 years of age; and finally, the 4th cycle, years 5 and 6, from 10 to 11 years of age (Rocha, 2017).

Instrument and Procedures

The Children's Chronotype Questionnaire (CCTQ), previously published in 2009 by Werner, LeBourgeois, Geiger et al., was adapted to French and administrated to population of Luxembourg. It is an instrument intended to be answered by the parents of children aged between 4 and 11 years. It consists of 27 items, based on the works that developed and used the Munich Chronotype Questionnaire and the Morningness and Eveningness Scale for Children by Carskadon, Vieira and Acebo (1993). A Portuguese version of the CCTQ was also developed (Couto, 2011; Couto, Gomes, de Azevedo et al., 2013) for native population. The CCTQ measures sleep-wake parameters: bedtime, time when the individual feels sleepy, sleep latency, wake-up time, get up time, and the time when the individual is fully awake, on days with schedules (school activities, for example) and on non-schedule days, i.e. free days, such as holidays or weekends.

The CCTQ has three chronotype measures represented in three scales. For each measure, there is a given calculation. In this paper, we only focus on one of the three measures/scales of the adapted questionnaire:

• Chronotype scale (CT) based on the measure of Question 27 (parents' knowledge of their children's chronotype). Parents were instructed on the concept and variability of the chronotype by reading an informative text. They then defined the child as to the chronotype using the options in Question 27, i.e. "More early morning than night type". The score varies from 1 (undoubtedly morning) and 5 (undoubtedly the evening type).

As for the protocol, the adapted French version of the Questionnaire was used after the authorizations of the authors of the original version (Werner et al., 2009) and of the authors of the Portuguese version (Couto, Gomes, de Azevedo et al., 2013) were obtained. First, an initial adaptation to the French language was made, followed by the mandatory back translations. In a first phase, the back translations were made from Portuguese into French, and a French version from the German questionnaire (original). Then, a final version was made, which was tested in a pilot group (n=35) in Luxembourg during the first quarter of 2016. The retest was made with the same group (n=35) three weeks later in the same quarter. After linguistic corrections, the questionnaire was formalized in its French version. To operationalize the study, this protocol involved the selection of students in Luxembourg schools (through authorizations requested to the Ministére de l’Éducation nationale, de l'Enfance et de la Jeunesse of Luxembourg, and representatives of schools and school groups).

The SPSS programme, version 23, was used to process and analyse the data. The means between the groups were compared using the univariate analysis of variance (ANOVA) tests and the Kruskal-Wallis non-parametric test in the analysis of cases with non-homogeneous samples regarding variance (verified with Levene's test). In the specific case of the Chronotype Scale (CT), the Spearman correlation coefficient was analysed following the procedure of the authors of the original version. We also used t- tests for independent samples in the cases comparing categorical variables means with two groups. However, in this paper we only report part of the data analysis conducted.

Part I - Results

As regards the results of the Pittsburgh Sleep Quality Index (PSQI) (n= 1109), the mean value of the overall PSQI is 4.14, which is significantly below the cut-off point of 5.00, t(1108)= -10.03, p <.001 indicating that, on average, children in this sample have good sleep quality. Compared to boys, girls report significantly better sleep latency and sleep duration (p=.001 and p <.001, respectively). There are no significant differences between girls and boys regarding overall PSQI. Among children in private schools and children in state schools, the latter have a significantly higher level of daytime dysfunction (p=.005). In general terms, the sleep quality (overall PSQI) of state school children is significantly worse than that of private school children (p=.026). The majority of the children (691) have a clear positive perception of the quality of their sleep (62.3%) compared to 418 children who manifested a poor perception about the quality of their sleep (37.7%).

The results of the Pittsburgh Sleep Quality Index (PSQI) indicate that children generally have a very positive perception of their sleep quality. However, these results do not compare with the Children's Sleep Habits Questionnaire (CSHQ-PT) relative data. This proves that the children in this sample have poor sleep quality.

For the Children's Sleep Habits Questionnaire (CSHQ-PT), only 883 complete responses were received, representing 77% of the total sample. In this instrument, the mean value of the Sleep Disturbance Index (SDI) is 46.12, which is significantly above the cut-off point of 44.00, t, indicating that these children generally have poor sleep quality. There are no significant differences between girls and boys regarding the SDI, with 46.9% of boys reporting sleep disturbances and 54.2% of girls also having sleep disturbances. However, boys wet the bed more often (p=.016). There is a significant difference regarding the level of daytime drowsiness (p=.018), girls wake up moodier (p=.011), have more difficulty getting out of bed in the morning (p=.019), and take longer to be well awake than boys (p=.004). In state schools, 179 children (41%) did not have sleep disturbances and 224 (50%) had sleep disturbances. In private schools, 257 children do not have sleep disturbances (58.9%), but 223 children do (49.9%).

Note . N = 883. M = mean; SD = standard deviation. Comparison between the 2 groups using the Mann-Whitney U (non-parametric) test. Cell with * = Statistically significant difference at the item level. Cell with ** = Statistically significant difference at the level of dimension.

¹ The value for each group includes 4 items: 5, 7, 8, e 21.

There are significant differences between children in state schools and children in private schools regarding 5 out of 8 dimensions. The SDI for state school children is significantly higher than that of private school children (p= .015). In contrast to private school children, state school children fall asleep more often in their parents'/brothers’ bed (p= .014), take longer to fall asleep (p= .016), are more afraid of sleeping in the dark (p= .035), wake up more often scared of nightmares (p= .046), snore loudly more frequently (p= .012), and wake up less often on their own in the morning (p= .011).

Part II - Results

For the Chronotype Scale (scale 3 of the Questionnaire - question 27), parents indicate more prominently that their children are more of the 'early' morning type than 'nocturnal' (18 parents indicate that children are of the early morning type versus 9 that say they are nocturnal). The morning type is the most perceived by parents regarding the sleeping and wakefulness behaviours of the children. Through a series of univariate analyses, the perception of the parents in relation to the chronotype of the children showed significant differences regarding the children’s real chronotype: F(7,479)= 5, p= .000, η2 .184. Later differences were found between the groups of parents pointing to the early-morning type and the remaining groups (i.e. 'more nocturnal than early-morning') and, for the real chronotype identified, the differences were between morning and evening types. It is more acceptable for parents to identify their children as being more of the morning type than of the evening type.

The sample of children presents, in an effective way, greater variability than expected with regard to the chronotype: 58 (31.5%) morning type, 48 (26.1%) intermediate, 66 (36%) evening type. Differently from the original study and the Portuguese validation study, the morningness measure (scale 3) was significantly associated (p <.05) with age. The scale referred to here is part of a complete three-scale context, and the first scale – Morningness/Eveningness - showed good validity (Cronbach's alpha: .70) after using the adapted French version.