Background
Childhood obesity is one of the most serious public health problems. The number of obese children under the age of five is gradually increasing all over the world. Forty-two million children under 5 years of age are estimated to be affected by overweightness and obesity worldwide [
1]. Overweightness and obesity in early childhood also lead to a higher risk of overweightness and obesity in adulthood [
2], and confer an increased risk of chronic inflammatory conditions including diabetes mellitus (DM), cardiovascular diseases, non-alcoholic fatty liver disease, and some cancers. Although cerebrovascular and cardiovascular events rarely occur in childhood, even in severe obesity, obesity in childhood is likely to result in a significant long-term economic burden on society, with associated excess lifetime health care [
3] and indirect costs [
4] including sick leave, reduced productivity, and premature mortality. Symptomatic pediatric lifestyle-related diseases are present in 5–15% of obese children and their incidence increases after late elementary school age. Moreover, obesity in childhood drives higher morbidity and mortality compared to obesity developed during adulthood. Therefore, intervention against childhood obesity is very important. This study focused on children aged 9 or 10 years who were subjected to screening and intervention for obesity to prevent adult and pediatric lifestyle diseases.
Amino acid (AA) profiles have been used as a biomarker of obesity and DM. We previously reported that the plasma concentrations of valine (Val), leucine (Leu), and isoleucine (Ile), as well as the total branched chain amino acids (BCAA), alanine (Ala), citrulline (Cit), and proline (Pro), were significantly higher in diabetic mice than in normal mice [
5]. Wang et al. [
6] reported a 12-year follow-up study showing that plasma levels of BCAA, tyrosine (Tyr), and phenylalanine (Phe) could be predictors of the future development of diabetes in nondiabetic subjects. Other studies have reported significant associations between the plasma levels of specific AAs and body mass index (BMI) [
7], AAs, and glucose regulation [
8]. In a study on Japanese obesity, Takashina et al. [
9] reported specific associations between specific AAs including Val, Leu, Ala, and Cit, the type/degree of obesity, and indices of glucose/insulin regulation in Japanese adults with normal glucose metabolism.
In this study, we analyzed the correlation of blood amino acids and obese metabolic states to assess whether plasma-free amino acid profiles can become useful biomarkers of lifestyle-related diseases in children with obesity. Moreover, we discussed the metabolic role of amino acids in children with obesity. This study included a clinical laboratory-based examination and measurement of the intima media thickness (IMT) of the internal carotid artery as a marker of metabolic state.
Methods
Study design
We retrospectively studied the medical records of 26 patients (male: 15, female: 11), aged 9 or 10 years, who presented in the Department of Pediatrics, Kumamoto University Hospital with moderate to severe obesity (defined as a degree of obesity ≥30%) at the first and second (after 6 months) screenings performed in Kumamoto City between April 2014 and March 2016. Degree of obesity was calculated according to the formula: ([real body weight–standard body weight depending on age] ÷ the standard weight × 100), as defined by the Japanese Society for Pediatric Endocrinology [
10].
Clinical evaluation
Clinical information, including age, sex, symptoms, present condition, medical history, medication use, and family history, was recorded on a standardized data form by the examining medical staff during the patients’ visits. The degree of obesity, body mass index (BMI), blood pressure, blood uric acid (UA), liver function [alanine aminotransferase (ALT), aspartate aminotransferase (AST), lactate dehydrogenase (LDH), and γ-glutamyltransferase (γ-GTP)], glucose tolerance [fasting blood glucose, insulin, C-peptide, and homeostasis model assessment-insulin resistance (HOMA – IR)], and blood lipid levels [total cholesterol (T-CHO), low-density lipoprotein cholesterol (LDL-CHO), high-density lipoprotein cholesterol (HDL-CHO), and triglyceride (TG)] were evaluated. The blood samples of obese children were collected after fasting for 12 h.
Assay of amino acid levels and measurement of intima media thickness
Plasma amino acids were analyzed using a liquid chromatograph mass spectrometer (SRL, Inc., Tokyo, Japan). The IMTs of the internal left and right carotid arteries were measured using an Aplio XG ultrasound machine (Toshiba Medical System Corporation, Tochigi, Japan) and double-checked by two technicians. The IMTs were measured at three points of both internal carotid arteries and averaged (Additional file
1).
Data quality analysis
Two researchers, who did not participate in the medical diagnosis,
ultrasonography, blood analysis, or medical record evaluation, performed the data and statistical analyses in this study.
Statistical analysis
We compared amino acid levels between obese children with and without impaired glucose tolerance using a t-test or Mann–Whitney U test. The factors IMT, LDL- and HDL-CHO (LDL/HDL ratio), amino acids, HOMA-IR, and UA were analyzed pairwise using Pearson’s correlation or Spearman’s rank correlation in IBM SPSS Statistics ver. 25. HOMA-IR and UA were dependent variables for predicting blood amino acid values, such as Val, Leu, and Ile, which are independent variables. The IMT and LDL/HDL ratio, were not dependent variables for blood amino acid values. A two-sided probability value of P < 0.05 was considered to be statistically significant. We also compared BMI, IMT, insulin, LDL-CHO, and amino acid levels before and after intervention using a paired t-test or Wilcoxon signed-rank test. Average values are presented as the mean ± one standard deviation.
This study was approved by the Ethics Committee of the Graduate School of Medical Sciences, Kumamoto University. Informed consent was obtained from the parents of all the children included in this study or the parents and children.
Discussion
Table
5 summarizes the relationship between blood amino acids and HOMA-IR, UA, LDL/HDL, and IMT. Obesity often transfers from early childhood through school age, and it extends into adulthood in an estimated 50% of cases. There are some reports that obesity, hyperlipidemia, and hyperglycemia in the adult are significantly correlated with IMT, and are risk factors for severely elevated IMT [
11,
12]. This correlation has also been seen in children [
13,
14]. Although arteriosclerosis is not common in obese children, their IMT tends to be higher than in non-obese children [
14,
15]. In this study, we found that IMT correlated negatively with HDL-CHO and positively with LDL/HDL ratio, although a positive correlation with LDL-CHO was not present. These findings suggest that obesity drives arteriosclerotic changes even in childhood. The risk factors for atherosclerosis include hypertension, hyperglycemia, and hyperlipidemia; however, few obese children develop hypertension. Hyperlipidemia and hyperglycemia are considered to be the most important risk factors for atherosclerosis. In our obese children group, blood glucose and insulin resistance did not affect IMT significantly, but the lipid metabolic parameter significantly correlated with IMT.
Table 5
Summary of correlations between blood amino acids and HOMA-IR, UA, LDL/HDL ratio, and IMT
The presence in childhood of a higher number of risk factors for the development of lifestyle-related diseases is associated with greater IMT in adults [
16,
17]. Raitakari et al. [
16] reported that a number of risk factors for atherosclerosis measured in 12- to 18-year-old adolescents, including high levels of LDL-CHO, BMI, and systolic blood pressure, were directly related to carotid IMT in adults. The presence of these risk factors at infant and school ages also affected IMT in adults. Therefore, we should consider treating children with obesity as a disease group, rather than simply as a group with lifestyle-related risk factors for future illness.
Elevated levels of amino acids such as BCAA, Ala, Glu, Asp, and Tyr, which are related to type II diabetes, have previously been shown in obese children with HOMA-IR ≥ 2.5 [
7]. We showed positive correlations between HOMA-IR and several amino acids, including TFAA, in children with impaired glucose tolerance (HOMA-IR ≥2.5). In these children, TFAA was also significantly correlated with blood glucose and insulin. When hyper-nutrition advances and impaired glucose tolerance develops, both glucose and amino acids accumulate. Cells such as hepatocytes and skeletal muscle cells become saturated, and this is considered to lead to hyper aminoacidemia.
Relevant associations between plasma amino acid levels and several other factors have also been documented. Insulin, growth hormone, glucagon, and IGF-1 play important roles in the regulation of energy metabolism in the living body [
18‐
20], and as demonstrated in this study, insulin affects plasma amino acid levels. Some reports have demonstrated a relationship between IMT and amino acids [
21,
22]. However, blood amino acids were not significantly correlated with IMT in our study. This phenomenon may be explained by a change in amino acid metabolism and insulin sensitivity in moderately obese children, because IMT was associated with LDL/HDL but not with blood insulin levels or HOMA-IR.
Associations have been shown between BCAAs and metabolic syndrome, obesity, type II diabetes, and/or insulin resistance [
7,
23,
24], and BCAAs are a cardiometabolic risk marker independently of BMI category [
25]. Increased plasma BCAA and lipids can lead to the development of β-cell dysfunction, which can accelerate the transition from an obese, insulin-resistant state to metabolic syndrome and type II diabetes [
24]. Pozefsky et al. [
26] suggested that impaired insulin activity and decreased utilization of amino acids in the muscles increased plasma BCAA levels owing to reduced uptake of BCAA in the muscles in lifestyle-related diseases. Moreover, Newgard [
24] reasoned that the increased circulating blood BCAA in obese and insulin-resistant subjects partly results from a decline of AA catabolism in their adipose tissue. Readily usable glucose and lipid substrates are considered to obviate the need for AA catabolism in adipose tissue by downregulation of the BCAA catabolic enzymes through the suppression of peroxisome proliferator-activated receptor-γ signaling in such metabolic adaptations.
Another particularly relevant amino acid is Ala. Würtz et al. [
27] reported that gluconeogenesis substrates including Ala increased in adults with impaired glucose tolerance. Moreover, Shimizu et al. [
28] reported that depletion of plasma Ala serves as a cue to increase plasma fibroblast growth factor 21 values and enhance liver-fat communication, resulting in the activation of lipolytic genes in adipose tissues.
Amino acids are not only essential nutrients serving as an energy source for the human body, but are also involved in many biochemical processes including the biosynthesis of purines and UA production. In recent years, many factors including BMI, alcohol intake, hyperlipidemia, and diabetes have been found to contribute to increasing blood UA levels [
29]. Our study indicated that in obese children, UA may be affected by amino acid metabolism rather than hyperglycemia and hyperinsulinemia. We found decreased levels of Gly and Ser with increased blood UA levels. Decreased blood Gly and Ser levels have previously been shown in adult patients with asymptomatic hyperuricemia or gout compared with healthy adult controls [
30]. The same study found increased blood levels of Ala, Val, Ile, and Orn in adult patients with asymptomatic hyperuricemia, but these amino acids were not correlated with UA in our study. It seems that Gly and Ser are related to the metabolic process of increasing blood UA level [
31]. Although Ser has no known relevant link with UA synthesis, Gly is needed for the de novo synthesis of purine [
32], which is the biosynthetic precursor of UA. More Gly may be consumed for the biosynthesis of purine in children with obesity under hyperinsulinemia.
This study has a number of notable limitations. Principally, our sample size was relatively small, in particular with regard to comparisons between children with and without impaired glucose tolerance. This limited the statistical power to draw firm conclusions. Finally, we did not evaluate the impact of dietary and lifestyle factors, or genetic factors including family history of obesity on amino acid patterns.
In recent years, early childhood obesity prevention is required because the prevalence of overweightness and obesity in children aged 5 years and below has been increasing worldwide [
33]. In Japan, examinations of physical and mental development were performed on young children at 1.5 and 3 years of age. Geserick et al. reported that most children who were obese between 2 and 6 years of age were obese in adolescence [
34]. In the future, we need to perform screening and intervention for obesity at the ages of 3 and 6, prior to their entry into kindergarten and primary school, respectively. This should involve assessment of their metabolic state. It would also be desirable to study junior high school students with obesity. Analysis of metabolic profiles including amino acids in obese children and adolescents from different age groups may reveal additional problems and remedies relevant to childhood obesity.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.