Effects of Trace Elements on Anthropometric Characteristics of Children: Cobalt and Childhood Body Mass Index

Objectives There are many reports on the effects of trace elements on human anthropometric characteristics. Among these elements, cobalt has consistently shown an inverse relationship with obesity risk. In the present study, we aimed to investigate the relationship between urinary levels of trace elements, focusing on cobalt, and childhood obesity, as indicated by the body mass index (BMI) in early adolescents, focusing on the participants' gender. Design A cross-sectional study was conducted in the Tokyo Teen Cohort study. Based on urinary samples, we obtained the anthropometric characteristics (weight and height) and potential covariates associated with childhood BMI for 1542 children (mean age=12 years; 860 boys and 682 girls). Methods Concentrations of urinary cobalt and 17 other trace elements were measured using inductively coupled plasma-mass spectrometry or inductively coupled plasma-atomic emission spectrometry. Results Pearson's correlation coefficient revealed an inverse relationship between the log of cobalt concentrations in the urine and the BMI for the boys (r=-0.125, p<0.001) and girls (r=-0.082, p=0.033). Multivariate analysis, adjusted for various covariates, reconfirmed the correlation between urine cobalt and the childhood BMI, only in the boys (beta=-0.14, p<0.001). Conclusions Among the 18 elements measured in the children's urine, cobalt may exhibit sufficient potency to decrease the risk of childhood obesity, particularly in boys. Future studies are required to clearly determine the magnitude of the effect and the underlying mechanism(s).


Introduction
Over the past three decades, the frequency of childhood obesity has been increasing, leading to a worrisome epidemic worldwide 1) .Currently, it is estimated that more than 38 million children under the age of 5 years and 340 million children/adolescents aged 5-19 years are overweight or obese 2) .
Owing to the risk of adulthood obesity 3) , cardiometabolic mortality, and morbidity 4) , childhood obesity is an important public health challenge 5,6) .Moreover, the Harvard Growth Study (1992) found that overweight adolescents had an increased risk of morbidity and mortality from coronary artery disease in the future, regardless of their adulthood weight 7) .
Childhood obesity is a complex problem with a multifactorial etiology, including environmental, genetic, and ecological factors 8,9) .For example, excess calorie intake in children may be an intrinsic consequence of unhealthy eating habits.This may include insufficient intake of necessary nutrients or excessive consumption of toxic substances.Additionally, trace elements may contribute to obesity by influencing metabolism.Many previous studies have examined trace element exposure as a risk factor for childhood obesity, and have reported inverse associations between cobalt concentrations in various biological samples and obesity risk (BMI) in children [10][11][12][13][14][15] and adults [15][16][17] . I addition, an experimental study demonstrated a difference in lipid profile (TG, HDL, and LDL) and body weight between mice exposed to cobalt and those that were not 18) .However, such relationships between cobalt and glucose and lipid metabolism have not yet been revealed in humans 19) .
Previous reports have suggested that the effects of trace metals may vary according to the child's gender, such as lowering of birth weight in male newborns due to elevation of arsenic or lead concentrations in maternal blood 20,21) and increase in the body weight of adult female participants due to higher hair cadmium levels 22) .Additionally, cobalt absorption and/or excretion can be influenced by gender, as serum and urine cobalt concentrations are higher in women than in men 23) .Similarly, studies in France 24) and Taiwan 25) have reported higher urinary cobalt (UCo) concentrations in women than in men.Furthermore, regarding the gender differences in the lipid profile after the onset of puberty, it would be important to assess the relationship between cobalt and childhood obesity in early adolescent boys and girls.
In the present study, we aimed to measure 18 trace elements in urine samples from children to assess their relationship with childhood BMI, mostly focusing on cobalt.Because gender plays an important role in cobalt biokinetics and affects chil-dren's anthropometric characteristics, we additionally compared cobalt concentrations, BMI, and their correlations between boys and girls.To our best knowledge, this study is the first to investigate the relationship between UCo and childhood BMI in early adolescents, focusing on the participants' gender.

Materials and methods
In the present retrospective, cross-sectional study, data and urinary samples were obtained from the Tokyo Teen Cohort (TTC) 26) .The TTC is a birth cohort study conducted by the Tokyo Metropolitan Institute of Medical Science for investigating children's physiological and psychological development, including self-regulation and personalized values on adolescents and their primary parents (usually mothers).In this communitybased survey, participants were recruited randomly from three municipalities in the Tokyo metropolitan area using the resident registry.Self-report questionnaires and interviews were conducted using 3171 children (10 years old at the baseline survey).In phase two of the study, the participants were aged 12 years when the data were collected for the current study.In this phase, urine samples were collected from 1582 children and stored at −80 °C until the metal analyses.

Analysis of urine sample
Concentrations of trace elements in children's urine samples were measured by inductively coupled plasma-mass spectrometry (ICP-MS) 13,36) or inductively coupled plasma-atomic emission spectrometry (ICP-AES), as previously reported 37) .ICP-MS (Agilent 8800, Agilent Technologies, California, USA) was used for determining Li, V, Cr, Co, Ni, Cu, Zn, As, Se, Sr, Mo, Cd, Ba, and Tl, and ICP-AES (Optima 2100, PerkinElmer, Massachusetts, USA) was used for measuring Na, Ca, Mg, and K.For the measurement preparation, urine samples were melted at room temperature and mixed with 0.5 % HNO 3 with 5-fold dilution in ICP-MS and analyzed by the multi-element standard solution XSTC-13 (SPEX CertiPrep, New Jersey, USA) as the external standard solution.For ICP-AES, the urine samples were diluted 10-fold with 0.5 % HNO 3 and analyzed using XSTC-2A (SPEX CertiPrep, New Jersey, USA) as the external standard solution.Measurements were repeated three times, and the average of the three measurements was used for statistical analyses.For instrument calibration throughout the measurements, at least 10 % of the analyses were external standards, and 5 % were blank (pure water).
For statistical analysis, the values under the limit of detection (LOD) were substituted with half the LOD.To correct for variations in urine dilution, the concentration of every trace element was expressed as a ratio to urinary creatinine concentration.To reduce the influence of outliers and normalize the right-skewed distribution, we used the natural logarithm of the urinary concentration of trace elements in the statistical analysis.Among the 18 trace elements, we focused on cobalt, since several studies showed an inverse association of this element with childhood obesity (Table 1).

Statistical analysis
Student's t-test (for continuous variables), Fisher's exact test (for categorical variables), or Mann-Whitney's U-test (ordinal scale) were used for comparisons between the two groups.Pearson's correlation coefficient was used to analyze the relationships between UCo and BMI.Multiple linear regression analysis was performed for assessing the relationships between UCo and urinary concentrations of the other 17 trace elements and BMI, controlling for possible confounding variables.All covariates were included using the forced-entry method.Analyses of all models were gender-stratified.The variance inflation factor (VIF) was employed for checking the multicollinearity problem among the variables.We used Bonferroni's correction to correct multiple comparisons, such as repeating the statistical tests 36 times (18 measured trace elements for both boys and girls), with a p-value<0.001(0.05/36), which was considered to indicate a statistically significant difference.All statistical analyses were conducted using IBM Statistical Package for Social Sciences (SPSS), version 27.0 (IBM Corp., New York, USA).

Results
The mean BMI of the children was 17.9 kg/m 2 and very close between boys and girls (17.8 and 17.9 kg/m 2 , respectively).The mean log UCo was −0.295 μg/g, and there was no significant difference between boys and girls (mean=−0.308and .The Student's t-test showed a significant difference in mean birth weight between boys and girls (mean±SD=3062±412 and 2995±401 g, respectively, p=0.002).The statistical analysis did not indicate any significant differences in the other characteristics between boys and girls (Table 2).
Pearson's correlation coefficient revealed a weak inverse correlation between log UCo and BMI in boys (r=−0.125,p<0.001) and girls (r=−0.082,p= 0.033) (Figure 2).Multiple linear regression analysis showed an inverse correlation between log  UCo and BMI in boys after adjustment for confounding factors (beta=−0.14,p<0.001) (Table 3).The VIF did not demonstrate a multicollinearity problem among the predictor variables (VIF<2.0).In addition, the statistical analysis failed to indicate any significant relationship between the BMI and the urine levels of the other trace element (Table 4).

Discussion
This study's findings showed that increasing UCo levels were associated with a decrease in BMI in boys.This association was confirmed after adjustment for several covariates, including genetic, behavioral, and environmental factors, in the multivariate analysis.Similarly, many previous studies have shown an inverse association between cobalt and obesity/overweight rates, regardless of age, gender, and type of the biological samples.Among them, five studies assessed only children [10][11][12][13][14] , one study had no age limitation 15) , and two studies assessed adults 16,17) .Consistent with the present study, Padilla (2010) 15) , Shao (2017) 11) , and Vrijheid (2020) 14) reported an inverse association between UCo levels and childhood BMI and weight.Błażewicz (2013) 10) showed lower plasma and blood cobalt concentrations in children with obesity than in those without obesity; however, the study failed to demonstrate the same results for UCo. Simlarly, Vigeh (2017) and Skalnaya (2018) reported higher levels of hair cobalt in children and adults with low body weight than in those with normal weight 13,17) . In adul women and children, an inverse correlation has been reported between cobalt levels in the toenail 16) and serum 12) , respectively, and BMI.In addition, experimental studies (mice and rats) have shown that blood/ urine/serum cobalt produce the same effects on animal weight/BMI 18,38) .Thus, the findings of the current study and previous epidemiological/exper-  imental studies suggest that cobalt may reduce the risk of obesity.Although many previous studies ignored the bioavailability of trace elements and their effects according to the subjects' gender, the present study examined UCo levels and stratified the effects by the participants' gender. Thee was no significant difference in the UCo levels between boys and girls in the present study. Howver, previous studies have reported higher levels of UCo in women than in men [23][24][25] . Thi difference may be induced by the greater iron demand in women.Cobalt and iron may share a common intestinal uptake mechanism 39) ; thus, iron deficiency (a common problem in young women) increases cobalt absorption and urinary excretion in animals 40) and humans 41) .Since the participants of the present study were in early adolescence (56 % of the girls did not experience menarche), the difference in UCo between girls and boys could not be detected at their ages.
Pearson's correlation analysis revealed an inverse correlation between log UCo and BMI in both boys and girls.When we adjusted for confounding factors, in the multiple linear regression analysis, a significant correlation was demonstrated only in the boys.These gender-related findings suggest that UCo is a protective factor against childhood obesity, predominantly in boys.The molecular or biochemical mechanisms underlying the reduction of BMI by cobalt have not been clearly understood.Tascilar (2011) found a correlation between plasma cobalt and the insulin resistance index (HOMA-IR), which suggests that cobalt acts as a regulator of glycogen depot by suppressing glucagon signaling and its effect on body weight 12) .In rats, cobalt administration was reported to result in decreased blood glucose levels, regulated glucose tolerance, and reduced body weight 38) .In addition, cobalt can decrease obesity risk by altering lipid metabolism, such as increasing leptin; the magnitude of the effect varies according to gender.For instance, leptin levels in women are higher after the onset of puberty 42,43) .Similarly, cobalt increases plasma HDL cholesterol and decreases LDL cholesterol, free fatty acids, and triglycerides 18) .Although cobalt may reduce childhood BMI by influencing lipid metabolism, we did not determine the level of leptin in the participants of the present study.Another possible underlying mechanism might be iron metabolism, as iron plays an important role during rapid growth periods, such as adolescence.
A recent study reported lower iron concentrations in children and adolescents who are overweight and a 50 % incidence of iron-deficiency anemia in individuals with a BMI above the 97th percentile 44) .Cobalt may influence iron metabolism, consequently increasing obesity risk by increasing the hemoglobin, hematocrit, and red blood cell counts in men 45) .Thus, cobalt may influence BMI by changing the metabolism of glucose, lipids, and iron differently in men and women.
In the present study, with a relatively large sample size, we considered several potentially confounding factors.However, the limitations of this study need to be addressed.First, the cross-sectional design of the present study may not draw conclusions regarding the causal relationship between UCo and childhood BMI.The generalizability of the results might be limited owing to the study design.Second, child anthropometric characteristics (i.e., weight and height) develop over several months to years; thus, a single UCo measurement may not reflect cumulative concentrations or exposure levels at an earlier life.Third, urinary excretion of cobalt is multiphasic, with a rapid increase in hours and a peak of elimination at 24 h following exposure 46) .In other words, the current study obtained the level of cobalt exposure via measurement of one spot-urine sample.It would be better to collect a 24 h urine sample for cobalt measurements.Finally, some considerable confounding factors associated with weight and BMI, such as the participants' diet and physical activities, were not adjusted for in the current study because information on these was limited in the TTC dataset.
In summary, among the 18 measured trace elements in the present study, only cobalt showed a significant inverse relationship with BMI in Japanese boys.Thus, cobalt may have sufficient potency to decrease the risk of obesity in children.Future epidemiological and experimental studies may need to clarify the magnitude of the effect and underlying mechanism(s).

Figure 1
Figure 1 Study flow chart

Figure 2
Figure 2 Pearson correlation coefficients between urinary cobalt concentration (UCo) and BMI in 860 boys and 682 girls

Table 1
Previous studies on relationship between trace elements, in different biological samples, and anthropometric characteristics

Table 2
Comparison of continuous and categorical variables between boys and girls a Data presented as mean ± SD or number (percentage) b Comparison between boys and girls Student's t-test was used for continuous variables, Fisher's exact probability test was used for categorical variables, Mann-Whitney's U-test was used for ordinal scale

Table 3
Relationships of log UCo and other variables to BMI: Multiple linear regression analysis by the forced-entry method Boys (adjusted R 2 =0.137)Girls (adjusted R 2 =0.096) a Standardized partial regression coefficient

Table 4
Relationships of 18 trace elements to BMI a Adjusted for age, birthweight, average sleep duration, father's BMI, mother's BMI, parental smoking, annual household income, father's education, and mother's education, by multiple linear regression analysis using the forced-entry method b Standardized partial regression coefficient