The relationship between dietary micronutrients intake and cognition test performance among school‐aged children in government‐owned primary schools in Kumasi metropolis, Ghana

Abstract Nutrients are critical for optimal brain development, and good nutritional status is associated with cognitive development and improvement. The relationship between micronutrients intake and cognition in Ghanaian school‐aged children has not been studied. The study investigated dietary intakes of micronutrients and cognition test performance of school‐aged children. A cross‐sectional study was undertaken among 438 school children, aged 9–13 years from ten randomly selected basic schools in Kumasi, Ghana. Socio‐demographic data were obtained from a structured questionnaire. Dietary intakes of iron, zinc, vitamin B6, folate, vitamin B12, and vitamin A were determined from repeated 24‐hr dietary recall data from 351 children, while cognition test was performed using a Raven's Coloured Progressive Matrices (RCPM), a 36‐question test. Among 351 children, 156 (44.4%) had inadequate zinc intake, whereas 96 (27.4%) had inadequate iron intake. More than 1 in 2 children had inadequate vitamin A intake while 55.8% and 53.0% had inadequate vitamin B12 and folate intakes, respectively. More school‐aged boys (66.3%) than girls (46.8%) had inadequate vitamin B12 intake (χ 2 = 13.393, p < .001), while for iron, folate, vitamin B6, zinc, and vitamin A, the differences were not significant. Mean RCPM test score differed significantly between school type (p < .001), but did not differ between the different ages, and between children with adequate and inadequate iron, zinc, vitamin B12, vitamin B6, and vitamin A intakes, except for folate intake (p = .050). Weak positive significant associations were observed between RCPM test score and zinc and folate intakes (p = .050). Dietary micronutrient intakes were inadequate in majority of these children, which put them at risk of weakened immune system and poor health, but did not show significant associations with RCPM performance. Further studies using other forms of cognition tests may help confirm our findings, and provide the impetus for the necessary interventions.


| INTRODUC TI ON
Proper brain function is a requirement for efficient cognitive function, and disruption of the brain reduces its efficiency (Bellisle, 2004). Good nutrition is necessary (Bryan, Osendarp, Hughes, & Baghurst, 2004), for brain function and cognitive performance, and thus should be adequately provided at all times (Bellisle, 2004). Poor nutritional status can indeed adversely affect brain function and cognitive performance (Bellisle, 2004). The proportion of children who do not obtain their full developmental potential is considerably large, and efforts to identify the specific causes of poor developmental outcomes are relevant (Peet et al., 2015). Micronutrient deficiencies are common among many developing countries, particularly among children, partly because of their higher physiological requirements and lower consumption of nutrient-rich foods (Thankachan et al., 2013). The consumption of monotonous diets that are low in animal products and rich in phytates is the potential causes of deficiencies in iron, zinc, vitamin A, and vitamin B 12 (Thankachan et al., 2013). New estimates show that nearly 250 million children under 5 years from low-and middle-income countries are at risk of not reaching their developmental potential, partly due to poor nutrition during pregnancy and the first 2 years of life (Black et al., 2017).
Therefore, poor micronutrients intake and nutrition in general, causing poor cognitive development, are among the reasons for many children not reaching their developmental potential (Figure 1).
Cognition is the ability to assimilate and process information received from different sources and to convert them into knowledge (Monti, Moulton, & Cohen, 2015). Cognitive processes include learning, attention, memory, language, reasoning, and decision-making (Monti et al., 2015). Children especially of school age, require proper cognition for higher and excellent academic performance (Yehuda, Rabinovitzm, & Mostofskym, 2006). Most cognitive processes and activities are highly associated with brain function, physiology, and structure (Gomez-Pinilla, 2008), and thus, its development may be affected by nutrition, physical activity, and social and economic status (Lemaire et al., 2010). During primary school education, deficiencies of vitamin A, iron, vitamin B 12 , vitamin B 6 , zinc, and folate among children can cause increased morbidity and negatively affect classroom attention and cognitive performance and thus devastate optimal schooling and academic performance (Fiorentino et al., 2017).
Vitamin B 12 is an important micronutrient found only in animal food sources and is required for complete brain development and cognition (Rathod, Kale, & Joshi, 2016). Several studies conducted in developing countries have reported a high prevalence of vitamin B 12 deficiency in children and attributed to inadequate vitamin B 12 intake (Finkelstein, Layden, & Stover, 2015;Swaminathan, Thomas, & Kurpad, 2015;Venkatramanan, Armata, Struppm, & Finkelstein, 2016). Vitamin B 12 deficiency is associated with poor cognitive development and growth in children (Strand et al., 2013;Moore et al., 2012;Van de Rest et al., 2012), and vitamin B 12 status in infancy have been associated with development and performance on neuropsychological tests after 5 years (Kvestad et al., 2017).
Zinc deficiency may affect cognitive development by causing changes in neuropsychological behavior, and motor development, through its interactions with other nutrients (Black, 2003). Interventional studies on effects of zinc supplementation and cognitive performance in schoolchildren have produced conflicting results. Randomized controlled trials in Chinese and Mexican-American children showed that zinc supplementation improved neuropsychological performance and reasoning, when compared with controls (Penland et al., 1999;Sanstead et al., 1998). In Canada, a randomized controlled trial found no significant effect of zinc supplementation on cognitive development in schoolaged children, compared with controls (Gibson et al., 1989).
Childhood iron deficiency anemia has been associated with delayed cognitive and motor development, and has negative effects on academic performance and educational fulfillment (Luo et al., 2017).
In school-aged children, cognitive function is impaired by iron deficiency with anemia (Grantham-McGregor & Ani, 2001). A randomized cross-over study by Sorensen et al. (2015) in Denmark found that school girls aged 8-11 years with few iron stores had a poorer overall "school performance" and poorer reading performance compared with girls with larger iron stores.
Vitamin A also regulates majority of the neurodevelopmental pathways that can plausibly influence cognition (Ali et al., 2017).
A study by Buckley et al. (2013) in rural Nepal found that vitamin A supplementation from preconception through postpartum did not improve cognition development of children at 10-13 years.
However, in Indonesia, provision of vitamin A supplement to infants showed slight improvement in motor development scores after 3 years (Humphrey et al., 1998).

| Study design and participants
A cross-sectional study design was adopted for this research. The study recruited 438 school-aged children between ages 9 and 13 years, living and attending basic schools within the Kumasi metropolis. Participants were recruited from ten (10) randomly selected basic government schools in the Kumasi Metropolis. In each school, all school children within the required ages in primary five were recruited for the study.

| Study area
The Kumasi Metropolis is one of the twenty-seven (27)

| Sample size and sampling procedure
Statistically, 10 schools were determined appropriate for the study and each of the schools was randomly selected from all governmentowned primary schools in the Metropolis, such that each school had an equal chance of being selected. However, a convenience sample size of 500 was used for this study and 50 pupils were to be selected from each school. In each school, all children in primary 5, who were within the ages 9-13 years were chosen for the study. All children who were therefore present on the day of data collection were included in this study. Since some of the schools had less than 50 in the chosen class, the eventual number was less than 500. Sick and physically challenged pupils were excluded from the study, since anthropometrics and physical fitness tests had to be performed and those in this category were not fit for those procedures.

| Data collection
Data were collected on their dietary micronutrients intake and cognition level. Socio-demographic data such as age, gender, and socioeconomic status of parent/guardians were obtained. Data were collected by researchers and other trained field assistants. Data were collected between September 2016 and May 2017.

| Assessment of dietary intake
A 24-hr triplicate recall on two weekdays and a weekend was used to collect dietary intakes of schoolchildren. Household food models of food items were used to identify quantities of foods eaten by participants. The weights (in grams) of foods consumed by participants were determined from the handy measures, and the composition of nutrients in meals was analyzed with the Nutrient

| Cognition assessment test
The Raven's Coloured Progressive Matrices (RCPM) test was used to assess the cognitive level of the schoolchildren. The cognition test was performed in a quiet environment, and the test procedures were explained to participants. The test contains three sets of twelve problems (36 colored questions), which measures fluid intelligence by problem-solving and abstract reasoning by analogy, and has been used extensively as a culturally fair test of intelligence (Raven, 2000).
The tests involved progressively geometrical designs and patterns with a missing piece, and each question has six to eight options to pick from and fill the missing piece. The cognitive tests were administered by well-trained research assistants. The children were given a booklet containing the test and answer sheets to select the correct answer for each question. This was explained to the pupil prior to the test.     In Table 5, we compare cognition scores by number of micronutrients that intakes were adequate for. Three groups were created, those with 0-2 nutrients adequate, those with 3-4 adequacy, and those with 5-6 nutrients adequate. Between these three groups, the mean cognition scores were not different (p = .753), although those with 5-6 nutrients adequate scored a slightly higher cognition score.

| RE SULTS
Similar observations were showed when the children were grouped into those with 0-3 versus those with 4-6 nutrients adequate.
There was weak, nearly significant positive correlation between cognition test score and dietary zinc and folate intakes (p = .05).
However, intakes of the other micronutrients had no significant association with cognition test scores (Table 6).

| D ISCUSS I ON
The present study reports the relationship between dietary intake of micronutrients and cognition test performance of school-aged children. There were more female participants than male participants TA B L E 1 Socio-demographic, micronutrients intake and percent cognition test score of school-age children (9-13 years) indicating that gender disparity in school enrollment among children in the Metropolis is probably not an issue any more.
Micronutrient deficiencies in children from developing countries remain a public health issue as they are persistently common. Deficiencies of micronutrients can be attributed to inadequate dietary intake, low bioavailability of micronutrients, and anti-nutritional inhibitors. Deficiencies in micronutrients increase risk of diseases and infection by weakening the immune system and further depleting nutrients stores (Katona and Katona-Apte, 2008). Many studies have pointed out that deficiencies in vitamin A and iron are among the causes of anemia, infection, low immunity, morbidity, and child mortality (Beard, 2001;Ramakrishan, Aburto, McCabe, & Martorell, 2004). This study reported poor micronutrients intake among the schoolchil- interventions are needed to avert problems that may arise especially for the girl children as they step into puberty, especially with regards to iron deficiency.
There were no statistical differences between dietary micronutrients intake and the different ages. However, the 10-year-old schoolchildren recorded the highest proportion with inadequate zinc and vitamin B 12 , while the nine-year-olds had the highest proportions with vitamin B 6 , folate, and vitamin A, implying that the lower ages of this physiologic group are likely to have poorer intakes.
Additionally, there were significant variations in micronutrients intake between the schools. The school with the highest proportions with adequate micronutrients intake showed close to the best cognition test scores compared with the other schools, implying that consumption of adequate micronutrients by school was related to better cognition test performance. However, a couple of schools that did not fall within the best of micronutrient intake also were recorded among the best cognitive performance. This suggests that micronutrients intake alone cannot be used to explain the cognition test performance.
Designed for children aged 5 through 13 years of age, the elderly, and mentally and physically impaired individuals, the RCPM test contains 36 progressively geometrical designs and patterns with a missing piece, listed in order of difficulty. It measures the test taker's reasoning ability, meaning making, and general intelligence (Raven, 1936). Twelve questions each are in set A and B, and another 12 questions in set AB, inserted between A and B. Our study revealed that majority of the schoolchildren (63.8%) passed the percent RCPM test score, scoring above average percent (50%), and the overall mean RCPM test score was above average (20.6 ± 7.7). As no national data exist for this average to be compared, it is not clear whether the average score by the children in this study is a good average or not, apart from the fact that close to two-thirds of the children scored over 50%.
Although in this study, the boys had a better RCPM test score than the girls, mean cognition test score did not differ by age, and whether the children had adequate or inadequate iron, zinc, vitamin B 6 , vitamin B 12, and vitamin A intake, except for folate. For folate, the children with adequate intake had a higher mean RCPM test score than those with inadequate intake, implying that folate intake, but not the other micronutrients, was related to the cognition test score. Evidence pointing to micronutrients intake and cognition is not consistent. A study by Boeke et al. (2013) in USA found no association between dietary vitamin B 12 and cognitive outcomes.
However, other studies, including Gewa et al., (2009) in Kenya and Ahmadi, Sohrabi, and Eftekhari (2009) in Iran, found that higher dietary vitamin B 12 intake among schoolchildren was associated with improved cognitive outcomes. Clearly, further studies are required to elucidate these relationships, and as more studies on this subject are conducted, better understanding will be reached.

| Limitations
The study assessed dietary intakes of participants but not serum levels, which may reflect the status of these nutrients better. It is also important to assess the overall patterns of intake rather than the specific nutrients. The dietary intakes were reported by schoolchildren directly, and therefore, over-or under-estimation of portion sizes is possible. The study however incorporated the use of visual portion estimates like household handy measures during the data collection to help participants recall food portions consumed and reduce bias. The authors also recognize that the 24-hr recall may not be the most reliable method for this age group. There was no comparison of income and other socioeconomic variables between the children, but it was expected that these children were from similar backgrounds by the fact that they were from government-owned primary schools.

| CON CLUS IONS
A large proportion of school-aged children 9-13 years, attending government primary schools in Kumasi Metropolis, had inadequate dietary micronutrients intake (with the poorest intakes within ages 9 and 10) and therefore may be at risk of deficiencies and their devastating consequences, such as weakened immune system, poor health, increased risk for infectious diseases, and delayed physical and neurodevelopmental. This is the first in the population we studied that tried to find the relationship between micronutrients intake and RCPM scores, and although with the exception of folate, most of the nutrients were not significantly related to this particu- Inadequate intake 215 20.5 ± 7.9 (0.5) Note: Data are presented as mean ± standard deviation (standard error mean), and p-value is significant at p ≤ .05. Post hoc analysis showed significant mean differences between school types with same alphabets (a, p-value = .016, b, p-value = .004, c, p-value = .002). † Independent t test. ‡ ANOVA.
TA B L E 5 Unadjusted mean comparison of combined adequacy for 6 micronutrients and cognition test score

ACK N OWLED G M ENTS
We acknowledge the staff of the selected basic schools in Kumasi Metropolis for the support during the course of this study.

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest regarding publication of this study.

E TH I C A L S TATEM ENT
Permission to carry out the study in the randomly selected schools was obtained from the Ghana Education Service (Kumasi Metropolis), following ethical approval from the Committee on Human Research Publication and Ethics (CHRPE) of the School of medical Sciences, KNUST. In each school, the study was explained to the Heads, who also gave approval and dates for data to be collected. All the children also gave informed assent to be included in the study, and parents/ guardians were followed up for a household survey. No blood samples were taken from participants. We declare that this manuscript is an authentic product of our research and is not published or communicated for publication elsewhere either in part or full. The manuscript is an accurate account of the study being reported.