Comparative Study of Inflammatory and Oxidative Stress Biomarkers in Different Metabolically Healthy Obesity Phenotypes

Aims: Obesity is the major contributor of the metabolic syndrome (MetS), but a unique phenotype of obesity known as metabolically healthy obese (MHO) shows healthier metabolic profile; however understanding of their biochemical correlates is poorly understood. Obesity is defined by Body mass index (BMI), but controversy exists regarding ethnic-specific BMI cut-offs. The present study used the Asian Indian BMI cut-offs to assess relationships of MHO phenotypes with oxidative stress (OS) and inflammation. Methods: In this case-control study, 299 metabolically-healthy (MH) respondents were divided into four groups as per Asian criteria for obesity: MH non-obese (MHNO), MH overweight (MHOW), MHO and MH severely obese (MHSO). Their oxidative stress and pro-inflammatory markers were measured. Results: Levels of hydroxyl radicals ( • OH), fluorescent oxidation products (FLOP), MDA, PCO and inflammatory markers CRP, TNF-α, IL-6 were highest in MHSO phenotype followed by the MHO, MHOW and MHNO groups (p > 0.0001), whereas antioxidant markers, CuZn-SOD, catalase, glutathione peroxidase and total antioxidant activity followed the reverse trend. The MHNO and MHOW groups showed significant difference with regard to ( • OH) radicals and FLOP. Moreover, • OH radicals, FLOP and inflammatory markers were significantly correlated to BMI in MHSO and MHO but not in MHNO and MHOW group. Conclusion: The MHO and MHSO phenotype display differences in terms of OS and inflammatory markers at lower BMI cut-offs, indicating that they may be on the way to becoming “unhealthy” obese. The lower BMI cut-offs proposed by Indian Consensus Group would help in understanding of manifestation of metabolic syndrome. of Inflammatory in Different Metabolically


Introduction
Obesity has been recognized as the major contributor to the global epidemic of metabolic syndrome, which has been defined as a cluster of conditions that occur together to increase the risk of heart disease, stroke and type 2 diabetes.
However, these metabolic abnormalities do not affect all obese people and the concept of "healthy obesity" was suggested by Sims several decades ago, in 1985, as a subtype in the classification of obesity [1].
The issue is under more active investigation now, and those who are obese but are not affected by metabolic disturbances have been designated as "metabolically healthy obese" (MHO) phenotype [1]. They are, by definition, insulin sensitive, have normal blood pressure, favorable lipid profile, a lower proportion of visceral fat, less liver fat and normal glucose metabolism despite having an excessive amount of body fat [2] [3], and are reported to be associated with substantially lower risk of metabolic complications [4] and account for about 10% -25% of obese people [5]. There is no definite classification criterion to describe MHO. However, the most acceptable criterion to define MHO in clinical practice is the absence of Metabolic Syndrome [3] [6], as defined by the NCEP-ATP III criteria [7] in overweight/obese subjects.
Biochemical differences have been suggested between "unhealthy obese" and "healthy obese". The former, but not the latter, have been reported to benefit from weight loss [8], and MHO subjects have been reported to have lower oxidative stress, inflammatory markers and diminished adipose tissue macrophage infiltration in comparison to metabolically unhealthy obese individuals [9].
The currently most accepted categorization of obesity and overweight is body mass index (BMI), which is used within each population to identify the proportion of people with a high risk of an undesirable health state that warrants public health or clinical intervention. WHO categorizes those with BMI ≤ 25 kg/m 2 as non-obese; 25 -30 kg/m 2 as overweight, and BMI > 30 kg/m 2 as obese. However, the last several years have witnessed a debate on whether ethnic-specific cut-off points for BMI for Asians are required in the light of scientific evidence that Asian populations have different associations between BMI, percentage of body fat, and health risks than European populations. In 2004, a WHO Expert Consultative Committee [10] [11].
In view of the foregoing, the present study has been undertaken to compare inflammatory and oxidative stress biomarkers in different metabolically healthy obesity phenotypes, which do not display the NCEP-ATP III risk factors for metabolic syndrome, and employing the BMI classifications for Asian Indians.

Subjects
A large number of healthy respondents attending outpatient departments (OPD) in government hospitals and pathologies at Allahabad, India, during the years 2015 to 2017 were screened for metabolic syndrome (MetS) risk factors described by the US (NCEP) ATP III [6], according to which respondents suffering from any three of the following risk factors, namely, Central obesity (waist circumference ≥ 102 cm/40 inches (male), ≥88 cm/35 inches (female)), Dyslipidemia (TG ≥ 150 mg/dl, HDL-C < 40 mg/dL (male), <50 mg/dL (female)) and Blood pressure ≥ 130/85 mmHg (or treated for hypertension) and Fasting plasma glucose ≥ 110 mg/dl) were excluded. Such respondents were included in the study and are termed "Metabolically Healthy (MH)" after ascertaining that they did not suffer from any infectious or non-communicable disease.
Their height and weight were measured following standard procedures, as recommended by WHO [12]. Their waist circumference (WC) was also measured, using a flexible measuring tape, at the natural waistline above the umbilicus and below the rib cage and recorded in centimeters. Body mass index (BMI) was computed as BMI = weight (kg)/[height (m) 2 ]. Their blood pressure was recorded following prescribed norms.
This case-control study was conducted on 299 respondents who fulfilled all inclusion and exclusion criteria. As described by guidelines by the Indian Consensus Group (for Asian Indians residing in India) [11], they were divided into the following four categories of metabolically healthy respondents: Respondents who gave their written informed consent were included in study.
The study protocol was approved by the Institutional Ethics Committee of Pop-

Blood Collection
Blood samples were collected, divided in sodium citrate anticoagulant-containing and plain vials, and processed to obtain packed red blood cells (RBCs), plasma and serum. RBCs were further processed to obtain hemolysate as described earlier [13], and stored at −80˚C until analysis.

Fasting Blood Glucose and Lipid Profile
The measurements of fasting blood glucose, total cholesterol, triglyceride and HDL cholesterol were performed with the autoanalyser kits manufactured by ERBA diagnostics Mannheim, Germany using semi-autoanalyser Chem-7, Erba Manheim. The LDL-Cholesterol was calculated by using the Friedewald formula: LDL-C = TC − HDL-C − (TG/5).

Non Enzymatic Total Antioxidant Capacity (TAC) by
Ferric Reducing Capacity of Plasma (FRAP) Assay FRAP was estimated by the protocol of Benzie and Strain [21].

Estimation of Cytokines
C-reactive protein (CRP) was estimated by Automated Bioanalyzer kits (Accurax Biomedical). Serum Inflammatory markers, Human IL-6 and TNF-α were estimated by ELISA kits (Elabscience) according to the manufacturer's protocol.

Statistical Analysis
Data were analyzed using Microsoft Excel 2010, Prism Graph Pad 5 and JASP 0.8 software. All results are presented as mean ± standard deviation. The statistical significance of the differences between groups was assessed using one way analysis of variance (ANOVA), followed by Tukey's honest significant difference post hoc test to assess all pairwise differences. Pearson correlation coefficients were obtained to see relationship between different variables. Unless stated otherwise, all values at 95% confidence with p < 0.05 were considered statistically significant.

Results
The baseline characteristics of the study population, presented in  Table 2.
The biochemical parameters including fasting plasma glucose, triglyceride, blood pressure were found to be significantly higher (p < 0.05), and HDL-cholesterol, lower in MHSO and MHO phenotype as compared to MHOW and MHNO, who did not differ from each other ( Table 2). Similar results were found for total cholesterol and LDL.
The concentrations of plasma/serum/erythrocytic oxidative stress markers and antioxidant markers were assessed in various phenotypes of obesity and again all OS markers, serum • OH radicals and FLOP, Erythrocytic MDA and PCO were significantly higher at p < 0.0001, and antioxidant enzymes, CuZn-SOD, CAT, and total antioxidant capacity (FRAP) low as obesity became more pronounced, based on one-way ANOVA. Obesity influenced plasma GPx statistically significantly at p = 0.032. One-way ANOVA also confirmed that inflammatory markers, CRP, TNF-α, and IL6, increased significantly at p < 0.0001 as obesity increased and followed the pattern MHSO > MHO > MHOW > MHNO.
Tukey's Honestly Significant Difference (HSD) post hoc test was performed to assess statistical significance of all pairwise differences between factor level means, and results are presented in Table 3.   When the results were presented graphically (Figure 1), it was clear that the difference between the metabolically healthy severely obese and the obese is more than that between any other groups with regard to all parameters namely hydroxyl radicals, FLOP and all inflammatory markers, and the difference between the non-obese and the overweight is minimal for all these parameters. Although the difference between MHNO and MHOW was not found to be statistically significant, the graphs indicate that all parameters for MHNO show a healthier trend than MHOW.
Since groups were categorized based on their BMI, relationship of selected OS markers, OH Radicals, FLOP, and inflammatory markers, CRP, IL6 and TNF-α were assessed by calculating Pearson's correlation coefficients and all correlations were found to be highly statistically significant. It was also important to know whether this relationship was observed in all obesity groups studied hence the correlations were separately computed for each group and results are presented in Table 4.
It was interesting to find that none of the parameters were significantly related to BMI in the MHNO and MHOW phenotype, and were maximally correlated to BMI in the MHSO, followed by the MHO phenotype, indicating the significance of BMI > 25 used here as the cut-off for MHO according to Asian Guidelines but is the WHO cut-off for overweight.

Discussion
Respondents were categorized as metabolically healthy because they had only a higher-than-normal BMI but no comorbidities. All risk factors for metabolic syndrome as prescribed by the NCEP-ATP III criteria were within the prescribed range. However, within this range, the fasting plasma glucose, triglyceride, blood   In a study designed to examine prevalence of the different metabolic phenotypes and to distinguish between unhealthy and healthy phenotypes of obesity [24], it was concluded that a healthy obese phenotype was associated with a better metabolic profile than observed in normal weight individuals with MetS, and increasing BMI had a significantly greater effect on estimates of liver fat and future CVD risk among those with MetS compared with participants without MetS.
The present study, conducted on respondents categorized into various obesity phenotypes based on Asian Guidelines [11], found that all OS markers, namely, Food and Nutrition Sciences markers, malondialdehyde and protein carbonylation, and total antioxidant capacity. When these enzymes decrease, the entire system moves towards increased oxidative stress, and in some cases of mild disturbance, the overall oxidative stress is not visible due to the ability of the system to restore equilibrium [13]. The same pattern is visible in the present study, where neither the erythrocytic MDA and PCO, nor the antioxidant enzymes, CuZn-SOD, catalase or GPX varies between the MHNO and MHOW groups, leading to similar total antioxidant capacity, a pattern not found in the obese and more severely obese phenotypes. The significant reduction in antioxidant enzymes and consequent reduction in the total antioxidant capacity in obese, and further in the severely obese are indicative of the breakdown of this adaptation.
• OH radicals and FLOP are the only two oxidative stress markers that show a small but significant difference between MHNO and MHOW groups. This may be attributed to their higher sensitivity. In this connection, Goossens [22] reported that MHO phenotype is relatively better protected against chronic diseases but MHO should not be regarded as a harmless condition, because they may have a higher propensity for cardio-metabolic disease. Plasma FLOP has been reported to independently predict the risk of subsequent cardiovascular disease events in epidemiologic studies [27], but has not been investigated for possible differences with regard to obesity phenotypes, especially MHO. This is observed in the present study. The protective profile of metabolically healthy obese postmenopausal women has been reported and is attributed to their lower concentrations of hepatic circulating alanine transaminase, reflecting lower hepatic insulin resistance and lower liver fat content [28]. In another study [29], branched-chain amino acids, aromatic amino acids and orosomucoid have been identified as serum biomarkers through metabolomics approach to distinguish physically inactive overweight/obese women with metabolic syndrome from metabolically healthy ones, and this is independent of body weight and fat mass.
We also found that inflammatory markers, CRP, TNF-α and IL-6 were significantly higher in the MHO phenotype and maximum in the MHSO phenotype, as compared to the MHNO and MHOW phenotypes, which did not differ from each other. Inflammatory markers are known to be related to adipose tissue distribution even in metabolically healthy phenotypes [22]. Lowered serum adipocytokines, TNF-α and adipocyte fatty acid binding protein (A-FABP) in metabolically healthy obese and non-obese and were significantly associated with unhealthy metabolic profile in non-obese Korean individuals [30] and evidence has been reviewed that suggests that normal adipose tissue function contributes to the healthy obese phenotype [8].
Overall, there was little difference between the metabolically healthy normal weight and overweight in the biochemical markers related to oxidative stress as well as inflammatory markers respondents, but the impact of obesity and severe obesity in metabolically healthy respondents was highly significant on oxidative stress parameters as well as inflammatory markers. The pattern of OS markers Further, it has been suggested that this healthier metabolic profile may not translate into a lower risk for mortality [9]. has been demonstrated that in vivo IL-6 release from gluteofemoral adipose tissue was markedly lower than from the corresponding abdominal subcutaneous fat depot both in men and women [31], suggesting that lower body fat may have a more beneficial inflammatory phenotype. These differences in disease risk are due to strikingly divergent functional properties of these adipose tissue depots, explaining the disparity seen in various metabolically healthy groups in the present study. The difference was seen in various metabolically healthy groups in the present study. Accumulation of adipose tissue in the upper body (abdominal region) is associated with the development of obesity-related comorbidities and even all-cause mortality.
In this connection, it must be remembered that the Asian cut-offs for BMI were used in this study [11] which classify a BMI of ≥23 kg/m 2 and ≥25 kg/m 2 as overweight and obese. Asians display a greater proportion of body fat for a given BMI than Caucasians [32] and are found to be susceptible to develop type 2 diabetes and coronary artery disease in spite of lower BMI and the use of universal BMI cut-off points to classify subjects as normal weight, overweight and obese are not predictive for metabolic risk factors. Asian Indians display exclusive features of obesity like pattern of body fat distribution in abdomen, central obesity, ectopic fat deposition particularly in liver, pancreas, muscles etc. are major contributory factors to driven metabolic syndrome and type 2 diabetes mellitus.
These metabolic perturbations are largely depending on combined effect of me- However, in the present study, BMI has been used for categorizing phenotypes and its appropriateness can be confirmed because significant correlation of all parameters with BMI is found in the MHO and MHSO groups but not in the MHNO and MHOW phenotypes.

Conclusion
Our study highlights that the metabolically healthy obese phenotype exhibits altered metabolic profile in terms of oxidative stress and inflammation at Asian Indian BMI cut-offs, making this phenotype at high risk for metabolic syndrome, type 2 diabetes mellitus and cardiovascular diseases. Efforts should be made to prevent obesity and severe obesity phenotypes, because systemic inflammation and oxidative stress are visible in these groups, which are not evident in the overweight phenotype. More studies are required using the Asian guidelines of BMI to manifest clustering of metabolic syndrome risk factors in comparison to Europeans.