The comprehensive assessment of the prevalence of metabolic syndrome in obese children and adolescents

,


Introduction
According to WHO estimates in 2016, more than 340 million children and adolescents aged 5-19 were overweight or obese (WHO, 2021).This rapid rise in obesity prevalence is attributed to increased consumption of high-energy foods high in fat and sugar, growing urbanization, and changing modes of transport (Jebeile et al., 2022).Childhood obesity is associated with a high risk of disability and premature mortality in adulthood (DeBoer, 2019).As the prevalence of obesity increases, the number of comorbidities associated with this condition also increase (Lim & Boster, 2023).However, not all obese people have an equivalent risk of cardiometabolic disorders such as hypertension, dyslipidemia, hyperglycemia, etc (Drozdz et al., 2021).Studies are showing that about 50% of obese adults are «metabolically healthy» if health is assessed by the absence of any criteria of the metabolic syndrome (MetS) and only 5% are «metabolically healthy» if any of the components of the MetS are absent and insulin sensitivity is normal (Smiths et al., 2019).In a previous study of clustering children with MetS, it was shown that 20% of children did not have insulin resistance (IR), although they had other cardiometabolic disorders such as dyslipidemia or hypertension (Aliusef et al., 2022).Although there is no single definition for a complication of obesity as metabolic syndrome in children (Reisinger et al., 2021).However, basically, all the definitions include at least three of the following criteria: obesity (central or abdominal), dyslipidemia, high blood pressure, impaired glucose tolerance, or insulin resistance (Aguilar-Gomez et al., 2020;Fernandez-Aparicio et al., 2021).

Aim
The aim of this study is to define the metabolic syndrome in obese children according to various known definitions and to identify a group that does not belong to the metabolic syndrome and which can be classified as «metabolically healthy obese» because it has one or two cardiometabolic disorders.In this article, we use a logistic regression model to consider which of the five definitions of diagnostic of MetS most affects the development of insulin resistance.

Materials and methods
The observational cross-sectional study on 82 children aged 12 to 17 with obesity was conducted at the Rheumocardiology Department of Children's Clinical Hospital No.6 in Kyiv.Inclusion criteria: alimentary obesity which body mass index (BMI)≥95 percentiles according to reference standards depending on age and gender (CDC, 2022).Exclusion criteria: patients with obesity, associated with genetic syndromes (Littleton et al., 2020).Metabolic syndrome (MetS) was established according to the criteria of five different sources: Adult Treatment Panel III (ATPIII) criteria modified for age (hereafter referred to as Cook et al.), Viner et el., de Ferranti et al., Ford et al., IDF Consensus Group (Aguilar-Gomez et al., 2020;Fernandez-Aparicio et al., 2021).
Table 1 summarizes five definitions of metabolic syndrome in children and adolescents.
Insulin resistance was determined by an update homeostatic model of insulin resistance (HOMA-  2).
The percentage prevalence of MetS found for the five definitions is shown in Table 3.
MetS is defined by the presence of 3 or more criteria, including obesity.In this study the prevalence of MetS among children with obesity was higher using the de Ferranti et al  defining MetS showed a significant difference in the prevalence of MetS between Viner et al. and de Ferranti et al. (Chisquare=11.29,p=0.023).No significant difference was found when other combinations of criteria were analyzed (p>0.05).
The prevalence of MetS criteria using different definitions is shown in Fig. 1 Using the logistic regression model the dependence of HOMA-2 IR> 2.26 (Y) and metabolic syndrome parameters (X1, X2…X5) according to five different definitions was established.ROC curves were plotted with all definitions (AUC greater than 0.5) (Table 4).The quality of the models was assessed according to the following classification: excellent (AUC≥0.9),very good (0.8≤AUC ≥0.9), good (0.7≤AUC ≥0.8), satisfactory (0.6≤AUC ≥0.

Discussion
In our study, there are more boys than girls among obese children.This trend is confirmed by a number of studies around the world.Thus, according to the 2019 Atlas of Childhood Obesity for children aged 10-19, the same trend was observed in 112 countries (Lobstein & Brinsden, 2019).The prevalence of MetS according to de Ferranti et al. can be explained by the widest range reference values of WC and lipid profile in puberty higher than 30 mU/l, and post-pubertal higher than 20 mU/l when hyperinsulinemia was diagnosed (Aguilar-Gomez et al., 2020).The mildpuberty period has the highest reference value for insulin, probably due to hormonal changes in this period.The fact that insulin resistance decreases with age was shown in our previous study, where a logistic regression model revealed the relationship between HOMA-2 IR and age (Aliusef et al., 2022) (AUC більше 0,5,p<0,05).Виявлено різну поширеність метаболічного синдрому залежно від різних підходів до діагностики.Побудова логістичної регресійної моделі показала, що параметри метаболічного синдрому впливають на ризик високого HOMA-2 за чотирма з п'яти критеріїв діагностики.
The quality of the model was assessed as «unsatisfactory» by de Ferranti et al. (Figure 3), «satisfactory» by Cook et al. (Figure 2) and Ford et al. (Figure 4), and «good» by Viner et al. (Figure5) and IDF (Figure6).

Figure 1 .
Figure 1.Number of MetS criteria (including obesity) according to different definitions

Table 2 .
Characteristics of the study group

Table 4 .
The characteristics of the ROC-curve five-factor logistic regression model of the dependence of HOMA-2 IR> 2.26 and different definitions of MetS