What is already known on this topic: Microminerals play key roles in brain development and function, but adolescents generally do not meet micromineral requirements due to poor dietary habits.

Nutrition plays a critical role in the development and functioning of the brain during the prenatal period and infancy (1), childhood and adolescence (2) and adulthood (3). Microminerals (also known as trace elements) include iron (Fe), iodine (I), zinc (Zn), copper (Cu), selenium (Se), and manganese (Mn). Of these, Zn is the second most abundant micromineral and is found in almost every tissue (4), and is involved in several metabolic functions including brain development and cognitive function (5, 6). Se is another essential micromineral which functions as an antioxidant and protects cellular membranes from oxidative stress. It is a major component of several functional proteins and enzymes, most importantly, the selenoproteins which are associated with thyroid hormones activation and brain functions (7, 8). Cu and Mn are also essential trace elements in the human body. Both function as enzymes cofactors, regulating normal metabolic functions (9). Both Cu and Mn are redox-active minerals and thus can work as antioxidants providing defense against oxidative stress and free radical damage. Both low and high Cu and Mn levels are known to negatively affect the IQ and cognitive functions in humans (10–12).

Deficiencies of microminerals are fairly common, and low levels of these minerals have been shown to adversely affect cognition and psychological behaviors (13). Micronutrient deficiencies in general are common in the Middle East, particularly in children, due to high consumption of fast foods and other unhealthy eating patterns. In Kuwait, poor nutrient intake and obesity among children are highly prevalent. However, to our knowledge, no data are available on the association of blood micromineral levels with cognitive function and academic performance in adolescents in Kuwait. We hypothesized a positive association between blood levels of microminerals (Cu, Mn, Se, and Zn) with cognitive function and academic performance in school in Kuwait. This hypothesis was tested in a nationally representative sample of adolescents.

Of the 1,422 students who returned the consent forms, blood samples were collected from 1,416 students. Micromineral data were available for 1,396 students, of which 681 were boys and 715 were girls. Mean (SD) age was 12.48 (0.93) years, ranging from 11 to 14 years. Socio-demographic characteristics of the study population are previously published (14, 15). Median (IQR) blood levels of microminerals (μg/dL) are shown in Table 1. Median Cu and Se levels in boys were significantly higher than girls (p < 0.05), whereas Mn and Zn levels were not significantly different between boys and girls. Figure 1 shows the distribution of SPM scores over the quartiles (Q1–Q4) of the microminerals. As shown, SPM scores were significantly higher in Q4 compared to the lower quartiles of Cu and Mn (p < 0.05). The difference in the SPM score across the quartiles of Se and Zn were marginally significant (p = 0.08 for Se and 0.06 for Zn). When the SPM scores of children with micromineral levels below the median were compared with those who had micromineral levels above the median, the mean difference between the groups (in favor of the higher group) was 5.3, 5.4, 4.8, and 4.8 points (p < 0.001 for each micromineral) for Cu, Mn, Se, and Zn, respectively (Figure 2). Table 2 shows the association between the microminerals and the SPM scores. As shown, all the microminerals were positively associated (p-values for all minerals <0.01) with the SPM score, whether the micromineral levels were used as continuous variable (Model 1) or as quartiles (Model 2). This positive association remained significant when the models were adjusted for confounding variables (Models 3 and 4, respectively). It is of interest to note that although the median differences across the quartiles of Se and Zn were marginally significant (Figure 1), after adjusting for other confounding variables (Model 4), the associations became highly significant. The coefficients (β) in the final adjusted model, when microminerals were used as quartiles (Model 4), were 2.03, 2.47, 1.64, and 1.88, respectively for Cu, Mn, Se, and Zn (p < 0.01 for all).

The association between microminerals and academic performance in mathematics is shown in Table 3. All micromineral tested (Cu, Mn, Se, and Zn) showed positive association (p < 0.001) with the percentage score in mathematics, whether the microminerals levels were used as continuous variable (Model 1) or as quartiles (Model 2). This positive association remained significant when the regression models were adjusted for confounding variables (models 3 and 5). We also tested the association between micromineral levels and academic performance in science and English and in the total score and got similar results (data not shown). Figure 3 shows the distribution of the total score across the quartiles of microminerals. While there was no statistically significant difference between Q1 and Q2, the higher quartiles (Q3 and Q4) were significantly different from the lower quartiles (p < 0.05). Similarly, the distribution of the percentage score in mathematics across the quartiles of microminerals is shown in Figure 4. The results are largely similar to the total score. On average, there was a seven-percentage point difference in the mathematics score between the Q1 and Q4 of the microminerals.

The data presented in this study demonstrates a strong positive association between the microminerals blood levels and cognitive function, as assessed by Raven’s Progressive Matrices, and academic performance, both in individual subjects like science and mathematics and the composite score in all subjects. On average, a seven-percentage point difference between the lower quartile (Q1) and the highest quartile (Q4) of the various microminerals tested was observed. These findings have significant public health implications, as micromineral deficiencies have been widely reported as a public health problem in various parts of the world. While this is a cross-sectional study from which a cause-and-effect relationship cannot be established, mechanistic data from the literature about microminerals and cognition make the positive association between the studied microminerals and cognitive function relevant. In the subsequent sections, we discuss the nutritional status of these minerals and the potential mechanisms by which these minerals affect cognition.

Zn deficiency is a public health problem worldwide (62). According to the world health organization (WHO), Food and Agriculture Organization (FAO) and International Zinc Nutrition Consultative Group (IZiNCG), 20% of the world population is at risk of developing Zn deficiency. An estimated 15–20% of the world population does not meet their RDA for Zn (63). Zn deficiency has been reported from many countries, in children, adolescents, and women of reproductive age (27, 64–69). Cu deficiency (based on plasma/serum Cu levels) has also been reported from several countries (70–72).

Limited data are available on the Cu and Zn status of Kuwaiti population. An earlier study in Kuwait on a relatively small sample size and on a rather broader age range (15–80 years old) reported that 36% of the general population is Cu deficient, whereas Zn deficiency was reported to be of no significant public health importance (73). Although the current study is based on a nationally representative sample of adolescents in Kuwait, we are unable to report the estimates of micromineral deficiencies in this population due to the absence of local references ranges for the whole blood micromineral levels.

The absence of established age-specific local cutoffs makes it difficult to determine the prevalence of deficiencies of these microminerals in adolescent population in Kuwait. This is further complicated by the variations in the specimen (whether plasma, serum or whole blood) used for analysis. While the values for some minerals are quite similar in serum/plasm and whole blood, it is significantly different for others. For example, Cu levels of serum and whole blood were reported to be quite similar, whereas serum Zn levels were significantly lower than whole blood Zn levels (74–76). Despite these issues, our data reveals that Cu and Zn levels in our population are significantly lower than the blood levels of these micromineral reported in the literature across various age groups from various countries (Supplementary Table 1). Median Cu level in our population was 23 μg/dL compared to the reported means (median or GM) ranging from 87 to 124 μg/dL. Similarly, median Zn level in population was 164 μg/dL, compared to the reported means (median or GM) ranging from 470 to 760 μg/dL. On the other hand, blood levels of Mn and Se in our study population were higher than all the studies reported (Supplementary Table 1). Adolescence growth spurt which increases the requirements for these minerals due to rapid growth, inadequate dietary intake due to overall poor dietary habits, higher levels of toxic heavy metals like Pb (14); and cadmium (unpublished data from this cohort) may increase the prevalence of these micromineral deficiencies.

This is the first systematically conducted study on a nationally representative sample that investigated the association between micromineral levels and cognitive function and academic performance. Another strength of this study is that we tested cognitive function using a language and culture-free test (Raven’s SPM) to minimize the influence of language and culture. SPM is the most widely used non-verbal test of cognition in subjects with a wide age range. Furthermore, we gathered data on academic performance on several subjects individually, like science, mathematics and English, as well as the composite academic score. We used a variety of statistical tools that show the robustness of the results. In addition, the regression models were adjusted for all the confounding factors that were associated with cognitive function and academic performance in this cohort. In addition, we employed the most sensitive method of mineral estimation (ICP-MS) with strict quality control measures. A limitation of this study is that a single measure of cognition (Raven’s SPM) was used which does not measure all cognitive abilities. In addition, dietary history and food intake were not evaluated as possible factors for micromineral status. The overall nutritional status (and dietary intake) is a known predictor of cognition, and it could be argued that the blood micromineral levels reflect the overall poor diet, which could have confounded the reported association between micromineral levels and cognitive function and academic performance. This hypothesis, however, does not seem plausible for two reasons. One, only a small proportion of the adolescents (<2%) were underweight, the remaining population was either normal, weight, overweight or obese. As such overall undernutrition is less likely to be a confounding factor. Second, we tested the association of several blood parameters that reflect nutritional status (including Fe, hemoglobin, vitamin B₁₂, folate and vitamin D) and none of these were associated with SPM score or academic performance.

In conclusion, this study highlights a very strong statistically significant positive association between micromineral levels and cognitive function and academic performance. Low cognitive function and academic performance during school age has life-long socio-economic implications. These findings have significant public health implications, as microminerals levels in adolescents are generally considered to be marginal. Further research is warranted to develop local age-specific cutoffs and to estimate the true prevalence of these micromineral deficiencies in the population.