Complete Blood Count-Derived Inflammation Indexes Are Useful in Predicting Metabolic Syndrome in Children and Adolescents with Severe Obesity

Background: Childhood obesity is a globally increasing pathological condition leading to long-term health issues such as cardiovascular diseases and metabolic syndrome (MetS). This study aimed to determine the clinical value of the Complete Blood Count-derived inflammation indexes Monocyte/HDL-C ratio (MHR), Lymphocyte/HDL-C ratio (LHR), Neutrophil/HDL-C ratio (NHR), and System Inflammation Response Index (SIRI) to predict the presence of metabolic syndrome and its association with cardiovascular risk markers (HOMA-IR, TG/HDL-C, and non-HDL-C) in children and adolescents with obesity. Methods: The study included a total of 552 children/adolescents with severe obesity (BMI: 36.4 [32.7–40.7] kg/m2; 219 males, 333 females; age: 14.8 [12.9−16.3] years), who were further subdivided based on the presence or absence of metabolic syndrome (MetS+ and MetS respectively). Results: The MHR, LHR, and NHR indexes (p < 0.0001), but not SIRI (p = 0.524), were significantly higher in the MetS+ compared to the MetS− subgroup, showing a positive correlation with the degree of MetS severity (p < 0.0001). Furthermore, MHR, LHR, and NHR were positively associated with cardiometabolic risk biomarkers (HOMA-IR: MHR p = 0.000, LHR p = 0.001, NHR p < 0.0001; TG/HDL-C: MHR, LHR, NHR p < 0.000; non-HDL-C: MHR, LHR p < 0.0001, NHR p = 0.000). Finally, the ROC curve analysis demonstrated that among the analyzed indexes, only MHR, LHR, and NHR had diagnostic value in distinguishing MetS patients among children and adolescents with obesity (MHR: AUC = 0.7045; LHR: AUC = 0.7205; NHR: AUC = 0.6934; p < 0.0001). Conclusions: In conclusion, the MHR, LHR, and NHR indexes, but not the SIRI index, can be considered useful tools for pediatricians to assess the risk of MetS and cardiometabolic diseases in children and adolescents with obesity and to develop multidisciplinary intervention strategies to counteract the widespread disease.


Introduction
Childhood obesity has emerged as a significant concern in developed countries, reaching epidemic proportions, with over 340 million children diagnosed as obese in 2020 [1].
Importantly, pediatric obesity significantly increases the long-term risk of developing metabolic disorders, including insulin resistance, dyslipidemia, hypertension, and glucose intolerance, collectively known as metabolic syndrome (MetS) [2][3][4].MetS is a serious multifaced condition that is on the rise among children and adolescents, mirroring the increasing prevalence of obesity [4][5][6].
Although the precise onset of the disease is still unknown, it is suggested that the buildup of free fatty acids in different tissues contributes to compromised insulin signaling and consequent insulin resistance [2,7].Additionally, studies indicate that inflammatory cytokines released from malfunctioning adipocytes, like monocyte chemoattractant protein-1 and tumor necrosis factor-alpha, stimulate macrophage migration to adipose tissues, amplifying cytokine production.Additionally, the decrease in adiponectin levels associated with obesity can worsen inflammatory responses in adipose tissues [2,7].
Recent research has shifted focus towards identifying early indicators of MetS and cardiovascular risk in children with obesity.Of particular interest are inflammatory markers derived from the complete blood count (CBC), which provide valuable insights into the systemic inflammation associated with obesity [11][12][13][14][15][16].
While most of the research utilizing these indexes has concentrated on adults [12,14,15,[17][18][19][20][21][22], to date there is a paucity of studies examining the correlation between childhood obesity and CBC-derived inflammation indexes in relation to MetS and cardiometabolic risks.
Therefore, the present study aimed to assess the utility of CBC-derived inflammatory indexes in stratifying and predicting outcomes, exploring their intricate relationships with childhood obesity, MetS, and cardiovascular risk factors.

Patients
In this retrospective study, a cohort of 552 children and adolescents diagnosed with severe obesity (219 males and 333 females) was examined.The median age (interquartile range) of the participants was 14.8 years (12.9-16.3),with a median BMI (body mass index) (interquartile range) of 36.4 (32.7-40.7).These individuals were admitted to the Division of Auxology, IRCCS Istituto Auxologico Italiano, Piancavallo-Verbania, Italy, for a 3-week multidisciplinary integrated body weight reduction program (BWRP).The program involved an integrated approach, including an energy-restricted diet, moderate aerobic exercise as part of physical rehabilitation, psychological counseling, and nutritional education [23,24].
Obesity was defined as the presence of a BMI ≥ 97th percentile for gender and chronological age by using the Italian growth charts [25].Exclusion criteria encompassed conditions such as renal and hepatic failure, secondary obesity (e.g., genetic obesity), acute infections, inflammatory, autoimmune, or malignant diseases, neurodegenerative disorders, as well as hematological and/or oncological conditions.
The diagnosis of MetS was based on the IDF (International Diabetes Federation) criteria for pediatric populations [26,27].
For individuals aged < 16 years:  This retrospective study was approved by the Ethical Committee of Istituto Auxologico Italiano, Milan, Italy (research code: 2022_03_15_05, acronym: REEOBEPED).All data included in the present analysis were collected from the medical records of patients hospitalized at our institution in the past.Informed assent (children/adolescents) and consent (parents) were obtained at the time of hospitalization for the use of clinical and laboratory data on the patient.

Anthropometric Measurements
Waist circumference (WC), height, and body weight were determined following the guidelines outlined in the Anthropometric Standardization Reference Manual [28].Standing height was measured using a stadiometer-equipped scale (Holtain Limited, Crymych, Dyfed, UK).Body weight was recorded to the nearest 0.1 kg utilizing an electronic scale (Ro WU 150, Wunder Sa.bi., Trezzo sull'Adda, Italy).Waist circumference was measured in a standing position with a non-elastic flexible tape measure, with participants gently exhaling midway between the lowest rib and the top of the iliac crest [29].

Laboratory Analyses and Metabolic Variables
Blood samples were collected using standard tubes after an overnight fast, followed by the determination of blood count and metabolic variables.Hematologic parameters were assessed using Beckman Coulter instruments.Leukocyte counts were conducted using the impedance-based method after erythrocyte (RBC) lysis, while leukocyte populations were identified based on volume, conductivity, and scatter properties using VCS technology.
Serum HDL-C and triglyceride levels were determined through colorimetric enzymatic assays from Roche Diagnostics, Monza, Italy.Serum insulin concentration was measured via a chemiluminescent immunometric assay employing a commercial kit (Elecsys Insulin, Roche Diagnostics, Monza, Italy), while serum glucose levels were determined using the glucose oxidase enzymatic method (Roche Diagnostics, Monza, Italy).All serum analyses for triglycerides, HDL-C, and insulin were conducted using the Roche Cobas 6000 analyzer.

Blood Pressure Measurement
Systolic (SBP) and diastolic (DBP) blood pressure were measured three times while the subject was seated and relaxed by using a sphygmomanometer equipped with a cuff of appropriate size placed on the right arm [30].

Statistical Analysis
GraphPad Prism 10.2.0 software for Windows (GraphPad Software, San Diego, CA, USA, https://www.graphpad.com/,(accessed on 8 August 2023)) was utilized for data analysis and visualization.The normal distribution and linearity of each variable were assessed using the Shapiro-Wilk normality test.As normality assumptions were not met, descriptive statistics for continuous and categorical variables in Tables 1-4, S1 and S2, and Figures 1-3 were presented as medians (interquartile range) or percentages.
Continuous variables were evaluated and compared among subgroups (obese MetS-/obese MetS+, degrees of MetS severity) using the non-parametric Mann-Whitney U test, while sex was considered a categorical variable.A two-way ANOVA was employed to explore the impact of sex on MetS prevalence.Fisher's exact test was utilized for comparing contingency tables and categorical variables.
The non-parametric Spearman's rank correlation test was employed to assess the correlation between CBC-derived inflammation indexes and the degrees of MetS severity, as well as the cardiometabolic risk factors.Binary logistic regression analysis was performed to determine the predictive value of CBC-derived inflammation indexes (considered independent variables) for MetS.
Receiver operating characteristic (ROC) curves were constructed to illustrate the ability to discriminate between MetS− and MetS+.Graphs depicting the raw data of CBC-derived inflammation indexes of MetS− and MetS+ were generated, and the area under the ROC curves (AUC) was calculated to evaluate the accuracy of these indexes.Optimal cut-off values for CBC-derived inflammation indexes, along with AUC values, sensitivity, and specificity for predicting MetS, were determined using the same data.
A significance level of two-tailed p-value ≤ 0.05 was adopted for all data analyses, with a 95% confidence interval applied to all datasets.Before patient selection, a power analysis was conducted.A sample size of at least 50 subjects for each group (MetS+ or MetS−) was determined to be adequate to detect a statistically significant mean difference of 0.004 for MHR, with a standard deviation of 0.07 (for both groups), using a t-test with a power of 80% and an α error of 0.05.
Despite the different prevalence of males and females in the population, gender was not a factor discriminating between the different distributions of MetS (p = 0.523).
The total white blood cell count, CBC-derived inflammation indexes, and cardiometabolic biomarkers in the whole study population and in the subgroups with and without metabolic syndrome (MetS+ and MetS−, respectively) are shown in Table 2.In the MetS+ subgroup, the total white blood cell count and all the subpopulations were comparable to that of the MetS− subgroup (Table 2).
Furthermore, the cardiometabolic indexes TG/HDL ratio, HOMA-IR, and non-HDL-C ratio were significantly higher in the MetS+ subgroup than in the MetS− subgroup (p < 0.0001) (Table 2).
The whole population was subdivided into three subgroups based on the number of altered criteria indicating the absence (or presence) of MetS (MetS 0-2, MetS 3, and MetS 4-5).

Discussion
The present study sheds light on the complex interplay between inflammation and metabolic syndrome (MetS) in pediatric obesity.Our analysis of children and adolescents with severe obesity unveiled significant associations between the CBC-inflammatory biomarkers MHR, LHR, and NHR and MetS, providing valuable insights into the underlying pathophysiology of cardiometabolic complications in this population.
MetS represents a cluster of cardiovascular risk factors associated with insulin resistance, primarily influenced by factors such as visceral obesity and systemic inflammation, which often manifest early in childhood and adolescence and worsen during adulthood [2,6].
However, evidence regarding these indicators and MetS primarily pertains to adult populations, with few conflicting findings regarding low-grade inflammation associated with pediatric MetS and its role in promoting cardiovascular disease.
For example, Turkkan et al. [35] reported that inflammatory hematological indexes, including NLR, PLR, MLR, and monocyte/HDL-cholesterol values, in adolescents with obesity were comparable to those recorded in normal-weight individuals.Mărginean et al. [11] also found comparable NLR values between children and adolescents with obesity and normal-weight individuals.In a study performed on adolescents with obesity, Dilek et al. [43] reported that PLR values were comparable between adolescents with obesity and healthy controls, while Yazaki et al. [44] detected no correlation between NLR and BMI in their study of children with obesity.In contrast, Aydin et al. [45] observed a significant increase in NLR and PLR values in adolescents with obesity compared to healthy controls.
The aforementioned disparities can be attributed to variations in study populations, including differences in sample sizes, age ranges, levels of obesity, gender distributions, and the presence or absence of MetS.
The present study aimed to assess the diagnostic value of CBC-derived inflammation indexes (MHR, LHR, NHR, and SIRI) in detecting the MetS presence in adolescents with severe obesity and their association with well-known cardiovascular risk biomarkers (HOMA-IR [46,47], non-HDL-C [48] and TG/HDL-C [49]).
The accurate assessment of MetS prevalence in children and adolescents with obesity remains challenging due to the lack of standardized criteria.This ambiguity has led to discrepancies in the reported MetS prevalence in different studies [50,51].
The prevalence of MetS among children with obesity varies widely in the literature, with reported rates ranging from 10% to 66%, influenced by the criteria used for the assessment [50,51].Consistent with these reports, our study found a MetS prevalence of 27.9% among adolescents with severe obesity.Specifically, according to the IDF criteria, the prevalence of obesity in childhood ranged from 16.4% to 44.2%, with a higher occurrence in males compared to females (31% males vs. 23% females in this study) [51,52].
In our study, a MetS prevalence of 26.4% was found among all patients with obesity, with waist circumference being the most frequently altered parameter, affecting 96.6% of the sample (no.patients: 533/552), followed by alterations in HDL-C levels detected in 47.6% of the patients (no. of patients: 263/552).Interestingly, a relatively low prevalence of increased triglycerides was found (9.2%, no.patients: 51/552), consistent with previous findings suggesting a low prevalence of this risk parameter in adolescents with MetS [50,53].The prevalence of elevated blood pressure (42.0%, no. of patients: 232/552) and fasting blood glucose levels (0.9%, no. of patients: 5/552) in our study closely matched those previously reported in the literature [4,12,50,[54][55][56][57].
In the present study, higher values of inflammatory indexes, particularly the monocytesto-HDL-C ratio (MHR), lymphocytes-to-HDL-C ratio (LHR), and neutrophils-to-HDL-C ratio (NHR), were detected in individuals with MetS compared to those without MetS, thus suggesting a potential role of inflammation in MetS development and progression among children/adolescents with obesity.
While it is widely recognized that obesity is frequently associated with persistent, low-grade systemic inflammation, particularly exacerbated in the presence of MetS [58], our research revealed that children/adolescents with severe obesity and MetS+ did not differ in terms of white blood cell (WBC) count or subpopulations compared to individuals with obesity without MetS.
By contrast, patients with MetS had significantly higher amounts of the monocyteto-HDL-C ratio (MHR), lymphocyte-to-HDL-C ratio, neutrophil-to-HDL-C ratio (NHR), thus highlighting the central role of HDL-C as an indicator of altered inflammation status.Notably, these parameters appeared to be primarily influenced by reduced HDL-C levels in MetS-positive individuals rather than by increased circulating leukocytes.
Spearman correlation analysis and binary logistic regression revealed a positive association between all the indexes (with the exception of SIRI) and the presence and severity of MetS.Furthermore, ROC curve analysis corroborated these findings by demonstrating significant predictive capability for LHR, MHR, and NHR, but not for SIRI, within our MetS+ population.
Systemic inflammation plays a crucial role in the pathogenesis of chronic diseases, serving as a key factor in initiating and exacerbating their progression [59][60][61].
Currently, there is a growing interest in assessing the potential significance of hematological pro-inflammatory markers for diagnosing and predicting various chronic conditions [62,63].
To investigate this aspect further, we assessed correlations between CBC-derived inflammation indexes and cardiovascular risk biomarkers (i.e., HOMA-IR, TG/HDL-C, and non-HDL-C).Our analysis revealed significant positive correlations between LHR, NHR, and MHR with all three cardiovascular biomarkers.These findings are consistent with previous research in adult populations, indicating that systemic inflammation indexes are associated with a higher risk of cardiovascular diseases [17,19,21,38,64].
Strengths and limitations should be carefully considered in our study.We have achieved an adequate sample size through meticulous power analysis, allowing for robust conclusions drawn from analyzing a great number of children/adolescents with severe obesity (no. of patients: 552) recruited in a single third-level center for the multidisciplinary treatment of obesity.Additionally, all the parameters were measured at the same central laboratory, thus eliminating the risk of variations between different laboratory settings.
However, it is important to underline some limitations.Firstly, the lack of a control group of normal-weight individuals (which was not the aim of the present study) prevented understanding the possible BMI-related differences between individuals with obesity and those with normal weight.Furthermore, the lack of longitudinal data prevented the establishment of the chronological sequence of MetS onset and development.Additionally, the inconsistent collection of information regarding the onset and duration of MetS treatment among participants represented an aspect that will need to be investigated with targeted future research.Despite our efforts to account for various potential confounding factors and considering the young age of the patients, we cannot definitively rule out the possibility of other variables influencing MetS that may affect the concentrations of CBC-derived inflammation indexes.Given the rapid physiological changes during puberty [65,66], including alterations in secondary sexual characteristics and hormonal profiles, the evaluation of these factors could be a goal for future research.
While our ROC curve analysis suggests the predictive utility of CBC-derived inflammation indexes for children and adolescents with obesity and metabolic syndrome, it should not be regarded as the sole definitive test to establish the clinical validity of these biomarkers.Exploring additional inflammation markers, such as cytokines and oxidation parameters, would be advantageous.Unfortunately, these measurements were not available in the recruited study group, but they can be taken into consideration for new prospective studies on the population of pediatric patients with obesity.Lastly, it is noteworthy that our study focused on children/adolescents with severe obesity within the Italian context.Therefore, our findings may not be directly applicable to populations with different degrees of obesity in different countries.
In conclusion, our study offers initial evidence supporting the reliability of the MHR, LHR, and NHR indexes in predicting MetS among children and adolescents with severe obesity.Moreover, the associations observed between these indexes and cardiometabolic risk factors seem to provide valuable insights for the precise assessment of MetS linked to obesity.Altogether, the results of the present study underscore the validity of the CBCderived inflammation indexes, which are routinely easily obtainable and at a low cost, in predicting MetS in pediatric subjects with obesity.Early identification of MetS in childhood and a deeper knowledge of the intricate relationships between childhood obesity, MetS, and cardiovascular risk factors can contribute to the development of more targeted nutritional and/or pharmacological interventions aimed at preventing the long-term cardiometabolic complications of pediatric severe obesity.
Presentation of the hematologic parameters, CBC-derived inflammation indexes, and cardiometabolic parameters.The p-value represents the comparison between the obese MetS− subgroup and the obese MetS+ subgroup calculated with the non-parametric Mann-Whitney U test.A p-value (p) ≤ 0.05 was considered statistically significant.Abbreviations: MetS-: children/adolescents with severe obesity without MetS; MetS+: children/adolescents with severe obesity and MetS; MHR: monocytes-to-HDL-C ratio; LHR: lymphocytes-to-HDL-C ratio; NHR: neutrophils-to-HDL-C ratio; SIRI: systemic immune response index; HOMA-IR: homeostasis model assessment-estimated insulin resistance; non-HDL-C: non-HDL cholesterol (total cholesterol-HDL-C); TG/HDL-C: triglycerides-to-HDL-C ratio.

Figure 1 .
Figure 1.Relationships between the degrees of MetS and the CBC-derived inflammation indexes in children and adolescents with severe obesity.Dot plots display median values of these indexes corresponding to MetS severity levels: MetS 0-2 (low grade, absence of MetS, in black), MetS 3 (minimal MetS, in light blue), and MetS 4-5 (full MetS, in dark blue).The difference between the MetS subgroups for each index was determined using the non-parametric Mann-Whitney U test.p ≤ 0.001; **** = p ≤ 0.0001; ns (not significant) = p > 0.05.Abbreviations: MetS: metabolic syndrome; MetS 0-2: patients with 0 to 2 altered criteria for MetS; MetS 3: patients with 3 altered criteria for MetS; MetS 4-5: patients with 4 to 5 altered criteria for MetS; MHR: monocytes-to-HDL-C ratio; LHR: lymphocytes-to-HDL-C ratio; NHR: neutrophils-to-HDL-C ratio; SIRI: systemic immune response index; r: correlation coefficient.Additionally, the MHR, LHR, and NHR indexes were positively and significantly correlated with the number of MetS altered criteria (Figure 2, p < 0.0001).

Figure 1 .
Figure 1.Relationships between the degrees of MetS and the CBC-derived inflammation indexes in children and adolescents with severe obesity.Dot plots display median values of these indexes corresponding to MetS severity levels: MetS 0-2 (low grade, absence of MetS, in black), MetS 3 (minimal MetS, in light blue), and MetS 4-5 (full MetS, in dark blue).The difference between the MetS subgroups for each index was determined using the non-parametric Mann-Whitney U test.p ≤ 0.001; **** = p ≤ 0.0001; ns (not significant) = p > 0.05.Abbreviations: MetS: metabolic syndrome; MetS 0-2: patients with 0 to 2 altered criteria for MetS; MetS 3: patients with 3 altered criteria for MetS; MetS 4-5: patients with 4 to 5 altered criteria for MetS; MHR: monocytes-to-HDL-C ratio; LHR: lymphocytes-to-HDL-C ratio; NHR: neutrophils-to-HDL-C ratio; SIRI: systemic immune response index; r: correlation coefficient.

Figure 2 .
Figure 2. Correlation between CBC-derived inflammation indexes and degrees of MetS severity.The non-parametric Spearman's rank correlation test demonstrated a positive correlation between CBCderived inflammation indexes and the degrees of MetS severity.Significance was defined as p ≤ 0.05 for all analyses with a 95% confidence interval.Abbreviations: MHR: monocytes-to-HDL-C ratio; LHR: lymphocytes-to-HDL-C ratio; NHR: neutrophils-to-HDL-C ratio; SIRI: systemic immune response index; r: correlation coefficient; p: p-value.

Figure 2 .
Figure 2. Correlation between CBC-derived inflammation indexes and degrees of MetS severity.The non-parametric Spearman's rank correlation test demonstrated a positive correlation between CBC-derived inflammation indexes and the degrees of MetS severity.Significance was defined as p ≤ 0.05 for all analyses with a 95% confidence interval.Abbreviations: MHR: monocytes-to-HDL-C ratio; LHR: lymphocytes-to-HDL-C ratio; NHR: neutrophils-to-HDL-C ratio; SIRI: systemic immune response index; r: correlation coefficient; p: p-value.

Table 3 .
Correlation analysis between CBC-derived inflammation indexes and cardiometabolic risk factors in the study population.

Table 4 .
Binary logistic regression analysis of the predicting variables for the presence of MetS.

Table 4 .
Binary logistic regression analysis of the predicting variables for the presence of MetS.