68-3

56/130

252Owing to the risk of adulthood obesity3), cardiomet-abolic mortality, and morbidity4), childhood obesity is an important public health challenge5, 6). More-over, 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 weight7).Childhood obesity is a complex problem with a multifactorial etiology, including environmental, genetic, and ecological factors8, 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. Addi-tionally, 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 children10-15) and adults15-17). In 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 not18). However, such relationships between cobalt and glucose and lipid metabolism have not yet been revealed in humans19).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 blood20, 21) and increase in the body weight of adult female participants due to higher hair cadmium levels22). Additionally, cobalt absorption and/or excretion can be influenced by gender, as serum and urine cobalt concentrations are higher in women than in men23). Similarly, studies in France24) and Taiwan25) 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 addition-ally compared cobalt concentrations, BMI, and their correlations between boys and girls. To our best knowledge, this study is the first to investi-gate the relationship between UCo and childhood BMI in early adolescents, focusing on the partici-pants’ gender.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 inves-tigating children’s physiological and psychological development, including self-regulation and person-alized values on adolescents and their primary parents (usually mothers). In this community- based survey, participants were recruited randomly from three municipalities in the Tokyo metropol-itan 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.We extracted data on body weight and height from the TTC dataset, as well as known potential covariates associated with childhood BMI, such as age (month), birth weight (g)27), sleep duration (weighted average of weekday and weekend sleep hours per night)28-30), parents’ BMI31-33), parental smoking34), household income, and parents’ educa-tion35). Forty participants’ data were excluded because of missing information regarding the chil-dren’s height and/or weight. Finally, we included 1542 children, 860 boys (55.8 %) and 682 girls (44.2 %), for statistical analysis (Figure 1). Analysis of urine sampleConcentrations 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 reported37). ICP-MS (Agilent 8800, Agilent Technologies, Cali-fornia, USA) was used for determining Li, V, Cr, Materials and methods