Correlation Between Nutritional Status, Cognitive Function and Daytime Sleepiness of Schoolchildren in Terengganu, Malaysia

This cross-sectional study aimed to determine the correlation between nutritional status, cognitive function and daytime sleepiness of schoolchildren in Kuala Nerus, Terengganu. It involved 85 schoolchildren aged 9 to 12 years. Nutritional status was determined through anthropometric measurements, biochemical analyses for hemoglobin, urinalysis and urinary iodine concentration (UIC). While cognitive performance was tested using Raven’s Colored Progressive Matrices (R-CPM). The daytime sleepiness of the respondents was assessed using the Pediatric Daytime Sleepiness Scale (PDSS). It was found that 16.5% of the respondents were stunted while 4.7%, 12.9% and 17.7% were obese, overweight and wasted, respectively. In addition, 37.7% of the respondents were anemic and the median UIC was 177.0 µg/l, which was categorized as normal for children. The majority (37.6%) of the respondents had an average cognitive level. Furthermore, the mean for daytime sleepiness based on PDSS score for overall respondents was categorized under an average level that was 14.1 score. In the present research, results found that all of the factors examined in this study were not correlated with respondents’ cognitive function. As the majority of the children had average to low levels of cognitive functions, it is recommended to investigate further factors associated with cognitive functions and, subsequently, to design and deliver appropriate intervention.


INTRODUCTION
Malaysian schoolchildren face malnutrition problems, including both underweight and overweight . Based on the latest National Health and Morbidity Survey (NHMS-2019), it was reported that 29.8% of children 5−17 years of age are overweight (15.0%) and obese (14.8%). A decade ago, the prevalence of overweight among children 0 to 13 years old was 5.3% (95% CI: 4.5−6.2), and of obesity 8.0% (95% CI: 7.5−8.6). Because of the greater age range, the NHMS-2019 reported a higher proportion. The prevalence of underweight, stunting and thinness among schoolchildren in rural areas was higher than among schoolchildren in urban areas (Chung 2015;Kuay et al. 2013).
Cognitive function among primary schoolchildren might be influenced by several factors, such as gender, birth weight, children's nutritional status, parental education, household income, hemoglobin and iron status (Hamid et al. 2011). The differences in nutritional status in children may influence their food intake. For example, it is common that children that are overweight and obese are triggered to increase their daily food intake. Mental functions have biological roots, and the impairment of some neurochemical processes owing to a lack of trace elements may have mental effects. Thyroid hormones are essential for brain development and proper brain function throughout life, and iodine is required for their production (Abel et al. 2017), and its deficiency can be associated with mental retardation (Janka 2019). Furthermore, it has been reported that iodine exerts neuroprotective effects and improves intellectual performance (Martínez García et al. 2018), which includes cognitive function. The ability to cognitively control impulses may play a role in determining nutritional status, as negative moods have resulted in an increase of food intake that may Rosli et al. lead to overweight and obesity among children (Tice et al. 2001). A study by Haile et al. (2016) found a significant correlation between heightfor-age with cognitive function among children. Height-for-age represents the accumulation of dietary deprivation over time, which may have an impact on children's academic achievement (Shariff et al. 2000).
Previous studies reported that children with low hemoglobin (less than 12 mg/dl) had a substantially lower cognitive function score than those with normal hemoglobin (more than 12 mg/dl) status as measured by Raven's test (Hamid et al. 2011). In terms of iodine intake, a previous study found that one third of the world's school-age population has urinary iodine below 100 µg/l, which indicates insufficient iodine intake, while 39.9% of school-age children in South East Asia also had insufficient iodine intake (Kuay et al. 2013). Furthermore, Kuay et al. (2013) also reported that the median Urinary Iodine Concentration (UIC) of children in urban and rural areas in Terengganu was 87.7 µg/l and 72.4 µg/l, respectively, which falls into the Iodine Deficiency (ID) range. Al-Mekhlafi et al. (2011) reported that children who suffered from iodine deficiency were more prone to having lower cognitive function and educational achievement than children with sufficient iodine intake.
In addition to nutritional status, previous research also suggests that daytime sleepiness can have a negative impact on students' academic performance and motivation (Abdulghani et al. 2012). Poor sleep was significantly correlated with lower performance of cognitive functions among schoolchildren (Lam et al. 2011). Sleepiness defined as individual self-reported total sleep time and daytime sleepiness are associated with poor academic achievement and other negative outcomes such as unstable mood (Drake et al. 2003). Further studies on daytime sleepiness are suggested as it may influence cognitive performance and, in particular, memory and cognitive function (Ahmad & Bashir 2017).
Therefore, based on the knowledge acquired from previous studies regarding the malnutrition problem among children in rural Malaysia and the present of associations between nutritional status, anemia and iodine deficiencies, daytime sleepiness with cognitive function this study aimed to assess the correlations of these factors among school children in Kuala Nerus.
Hence, it can provide the latest data on factors that associated with cognitive function among schoolchildren aged 9 to 12 years old as well as baseline information regarding daytime sleepiness and cognitive function among this demography which is not widely reported in previous studies especially in Malaysia. In addition, the data gained from this study can also be used as a baseline for schoolchildren aged 9 to 12 years old in other states other than Terengganu in similar situations.

Design, location, and time
The present study was a cross-sectional study carried out in two primary schools located in Kuala Nerus, Terengganu. Two primary schools were chosen, namely, SRI Al-Amin and SK Tanjung Gelam. The respondents were sampled in both schools within a two-month period. Ethical approval for this study was obtained from the Research, Management and Innovation Centre, Universiti Malaysia Terengganu with reference number: UMT/RMIC/2-2/62 (81).

Sampling
In the present study, convenience sampling was used. There were five primary schools that had been selected in Kuala Nerus. However, through convenience sampling, only two primary schools were chosen to be involved in this study. These schools were chosen because of easy accessibility as their headmasters were willing to cooperate by giving their permission. Besides, the respondents were available during the data collection.
The criterion for inclusion of the respondents in both schools were schoolchildren aged 9 to 12 years who were approved by their parents or guardians. This age group was chosen based on a discussion with the school administration that allowed researchers to be involved with at least 9 years of children. Information consent sheets that thoroughly explain the objectives and methodology of this study were also given to the respondents and parents along with the consent sheet. The Yamane (1967) formula was applied in order to determine the sample size of this study with a 95% confidence interval. The common margin of error used in other studies ranges from 5% to 10%. Therefore, the margin of error used Nutritional status, cognitive function and daytime sleepiness of schoolchildren in this study was in between the range, which was 7%. The size of the population of this study was 145 respondents, and on the basis of the formula, the calculated sample size was 85 respondents.

Data collection
The tools used for sampling in the present study were a questionnaire, anthropometry measurements and biochemical analyses. The questionnaire was divided into three sections: Section A contained socio-demographic information, while Section B included eight items assessing respondents' daytime sleepiness following Drake et al. (2003), whereby a higher score on the Pediatric Daytime Sleepiness Scale (PDSS) indicated greater levels of sleepiness. Lastly, Section C was the cognitive function test, with 36 matrices of diagrammatic puzzles of Raven's Colored Progressive Matrices (R-CPM) (Raven 2008).
Anthropometry data used for this study included body height, body weight and BMI. In this study, weight was measured using a Tanita digital body fat monitor/scale UM-026 (Tanita, UK). Body height measurements were also taken without shoes and socks (SECA, Germany). The weight and the height were measured to the nearest 0.1 kg and 0.1 cm, respectively (Hamid et al. 2011). Height-for-age was then categorized as tall, normal, stunted, and severe stunted while BMI-for-age was classified as obesity, overweight, normal, thin, and severe thin (WHO 2007). The WHO AnthroPlus version 1.0.3 was used to determine Z-scores for height-for-age and BMI-for-age among children.
In terms of biochemical analysis, the instrument used for testing hemoglobin concentration was a photometer called HemoCue ® . The urinalysis was performed by using the Urine Analyzer PocketChem™ UA PU-4010 in order to determine the presence of bilirubin, blood, glucose, ketone, leukocyte, nitrite, pH, protein, specific gravity and urobilinogen. Only a small amount of the mid-stream of urine was needed for the test. The laboratory procedure widely used to determine the concentration of urinary iodine is based on the Sandell-Kolthoff reaction (Jooste & Strydom 2010).

Data analysis
In the present study, the Statistical Package for Social Sciences (SPSS version 20.0) was used to analyze the data. Spearman's Rank Order Correlation was used for non-parametric test to measure the relationship between the parameters of the test. In all analyses, p<0.05 was used to indicate significant outcomes. Table 1 shows the socio-demographic characteristics of children. The majority (45.9%) of respondents were aged 11 years old, all from Malay families. More than half of the children live in households that consist of four to six people. In terms of parents' employment status, most of their fathers (92.9%) were working and more than half of mothers (67.1%) were also working. Table 2 presents the mean height, body weight, BMI-for-age and height-for-age of the children. In terms of BMI-for-age, the majority of the respondents (56.5%) were in normal condition, while 4.7%, 12.9%, 10.6%, and 7.1% were obese, overweight, wasted and severely wasted, respectively. On the other hands, majority (83.5%) of the respondents had normal heightfor-age while 9.4% and 7.1% were stunted and severely stunted, respectively.

Nutritional status based on anthropometric measurement
As stunting reflects past nutrition, during their early childhood, some of these children might have experienced poor food intake and had some sort of infection ). In addition, they probably were consistently living in an unchanged situation due to underprivileged living conditions (Shariff et al. 2000). Picauly & Toy (2013) reported that the malnutrition conditions, such as stunting, experienced by school-age children will affect their ability to learn in school and, subsequently, may affect their learning achievement. Besides, it has been reported that children's sex and age affect both wasted and stunted. With increasing age, they are more prone to stunting but without wasting (Shariff et al. 2000).

Nutritional status based on biochemical analysis
In the present study, the majority (62.35%) of the respondents had a normal reading of hemoglobin concentration, while more than one third of the respondents were anemic. Ngui et al. In the present study, the median UIC for children in Terengganu (177.0 µg/l) was higher compared with the other regions, as stated above. The reason for acceptable UIC among schoolchildren in Terengganu might be due to the availability of iodine rich food sources to (2012) reported that anemia among adolescent is caused by physiological iron losses during menstrual bleeding in female and increased iron needs during the growth spur period.
Based on Table 3, the overall median UIC of respondents was 177.0 µg/l, which was categorized as a normal UIC for children. In general, this reflects that the children had sufficient iodine intake in their daily food intake. According to the study by Selamat et al. (2010) in different locations, the median UIC of children aged eight to ten years was in the normal range, in terms of Malaysian data (109 µg/l), and data for Peninsular Malaysia (104 µg/l), Sabah (150 µg/l) Nutritional status, cognitive function and daytime sleepiness of schoolchildren residents of the community, resulting in a higher frequency of iodine food consumption and a lower risk of iodine deficiency. A study by Kuay et al. (2013) showed that the median UIC of children who lived in urban areas in Terengganu was 87.7 µg/l, which falls into the iodine deficiency range. However, in the present study, the median UIC of children who lived in Terengganu was categorized as a normal UIC for children. Thus, the consumption of iodine among children in this state has improved over the last five-year period.
In terms of urinalysis, all of the respondents had no glucose in their urine. The majority (98.82%) of respondents had no protein in their urine. According to Simmerville et al. (2005), protein is not normally found in urine, as most proteins are too big to diffuse out of the glomerular filtrate. All of the respondents had normal urine pH, which was between 4.5 to 8.0 and this indicates that the respondents did not have UTI or calculi. Furthermore, the majority of the respondents (76.47%) had normal Urinary Specific Gravity (USG) and for some respondents that showed high USG, they might be suffering from dehydration due to inadequate fluid intake in school. Some of the factors that raise specific gravity include dehydration, diarrhea, emesis, UTI and proteinuria as well as water restriction (Simmerville et al. 2005). However, 16.47% of the respondents showed positive blood results in their urine. The abnormal blood found in urine among male respondents might be due to infection, whereas positive blood results among female respondents were probably because they were undergoing menstruation when the test was carried out. Lastly, the majority (96.47%) of respondents had negative leukocyte esterase in their urine, while those who showed positive results were suspected of having an infection.

Raven cognitive function
As shown in Table 4, the majority (37.6%) of the respondents had an average cognitive level, followed by high average, low average and extremely low cognitive levels with 25.9%, 12.9% and 10.6%, respectively. The mean R-CPM score for overall respondents was 93.5, in the range of 9.0 to 135.0 of total scores. The R-CPM test depends on non-verbal logical reasoning.
In addition, a recent study reported that varied improvements in cognitive skills (as  measured using R-CPM) were observed for different variables in both the intervention and control groups of preschool children with no robust evidence for physical activity-interventionrelated improvements (Jaksic et al. 2020).

Daytime sleepiness
Daytime sleepiness, including sleep continuity, how rested an individual feels upon waking up, and how alert one is during the day, has an important effect on educational outcomes in children (Blunden et al. 2018). In addition, sleep can also affect young adolescents' mental, social, physical and psychological components (Blunden et al. 2018). However, lack of sleep time and daytime sleepiness among schoolaged children were linked with low scores in neuropsychological attention studies . A meta-analysis investigating the effects of school start times on sleep and daytime sleepiness has discovered that later school start times result in longer sleep durations and reduced daytime sleepiness in adolescents (Bowers & Moyer 2017). In a systematic review, Meyer et al. (2017) found a moderate use of the Paediatric Daytime Sleepiness Scale (PDSS) to assess daytime sleepiness. This tool provides for the tracking of factors that influence children's and adolescents' excessive daytime drowsiness. The mean score for daytime sleepiness based on the (PDSS) for overall respondents was 14.1 with a range of 5 to 25 scores. It was categorized under average level scores. On the other hand, higher PDSS scores indicate a higher level of sleepiness associated with the individual's total sleep time (Drake et al. 2003). Based on a study by Drake et al. (2003), the mean score PDSS for overall respondents was 15.3, almost similar to the mean score of the present study, which was at 14.1. Another study by Perez-Chada et al. (2007) reported that the PDSS (mean score: 15.7) was robustly and significantly associated with academic failure.

Correlation between nutritional status and daytime sleepiness with cognitive function among respondents
The present study found no significant correlation between height-for-age and BMIfor-age with cognitive function. The results also demonstrated no significant correlations between hemoglobin and urinary iodine concentration with cognitive function. Lastly, there was also no significant correlation between daytime sleepiness and cognitive function, as shown in Table 5.
These findings are in agreement with a study by Hamid et al. (2011) that reported heightfor-age z-scores are not apparently correlated with cognitive tests. In contrast, Haile et al. (2016) found a significant correlation (r=0.38; p=0.001) between children's height-for-age and cognitive function. Another study also reported that brain functions including cognition, memory and locomotors skills are influenced by undernutrition (Ranade et al. 2008). A study by Haile  et al. (2016) reported no significant correlation between BMI-for-age with cognition, similar to the present findings.
In contrast, a study by Hamid et al. (2011) reported that hemoglobin concentration was significantly correlated with cognitive function (r=0.19; p=0.006). In their findings, it was reported that children with iron deficiency had significantly lower cognitive function test scores than their normal hemoglobin counterparts (Hamid et al. 2011). Another study by Al-Mekhlafi et al. (2011) among children in rural areas revealed that those suffering from iron deficiency anemia were more prone to having lower cognitive function and academic performance than healthy children. Across past studies, it has been conclusively reported that iron deficiency anemia delays the psychomotor development and impairs the cognitive scores of pre-school and school-age children in Thailand, Indonesia and Zaire (Al-Mekhlafi et al. 2011). Several studies have linked iron deficiency anemia to cognitive impairment, psychomotor disorders, and behavioral problems in children of all ages. Disorders of attention, emotion, intelligence, and sensory perception are the most common cognitive functions impaired by iron deficiency. These long-term impacts can have a negative impact on learning ability and professional ability acquisition (Radlowski & Johnson 2013;Jáuregui-Lobera 2014).
Despite the well-documented importance of adequate iodine intake for early-life neurodevelopment, Kippler et al. (2016) found no association between the mothers' urinary iodine concentration (median 172 µg/l) and children's general cognitive score, even when comparing children born to mothers with an inadequate iodine status (urinary concentrations less than 150 µg/l) (36% of the women) to those born to mothers with an adequate iodine status (urinary iodine concentration of more than 150 µg/l). This might be due to the iodine deficit that was discovered was not be frequent or severe enough to impact neurodevelopment.
The present study also shows no correlation between daytime sleepiness and cognitive function. This finding also was not in agreement with Lam et al. (2011), who reported that daytime sleepiness was significantly correlated with cognitive function among schoolchildren. Another study by Moreau et al. (2013) also reported that lack of sleeping time among school-aged children is associated with lower performance on neuropsychological tests of attention. Drake et al. (2003) stated that sleepiness would be related to individual total sleep time and quality that would be associated with poor academic achievement and other negative outcomes, such as poor mood.

CONCLUSION
In conclusion, the majority of the respondents had normal growth for both the indicators of height-for-age and BMI-forage. Based on urinary iodine analysis, the results indicate that they received sufficient micronutrients such as iodine in their daily food intake. The majority of them had average cognitive function levels. In the present study, there was no significant correlation (p>0.05) between anthropometric indices (height-for-age and BMI-for-age), hemoglobin concentration, urinary iodine concentration and daytime sleepiness with children's cognitive function.
The present study provides the latest data about the correlation between nutritional status, daytime sleepiness and cognitive function among schoolchildren in Kuala Nerus. This data is useful for future reference, especially the data about daytime sleepiness, as it was not widely investigated in the previous studies especially in Malaysia. As the majority of the children had average to low levels of cognitive functions, it is recommended to investigate further factors associated with cognitive functions and, subsequently, to design and deliver appropriate intervention.