Serum Adipocytokines Levels and Their Association with Insulin Sensitivity in Morbidly Obese Individuals Undergoing Bariatric Surgery

Abstract

Background

Obese adipose tissue secretes a variety of adipocytokines that act as metabolic regulators with complex mechanisms. Our objective was to compare serum concentration of a panel of adipocytokines between obese and non-obese individuals and identify any distinct patterns correlating with insulin sensitivity in obesity.

Methods

We designed a cross-sectional study among obese (body mass index [BMI] ≥30 kg/m², n=62) and non-obese (BMI <25 kg/m², n=32) individuals to compare circulating levels of the adipokines, such as adiponectin and resistin in conjunction with the measurement of the levels of inflammatory cytokines including C-reactive protein (CRP), interleukin (IL)-6, IL-8, monocyte chemoattractant protein (MCP)-1, and tumor necrosis factor (TNF)-α using Luminex multiplex immunoassay with drop array technology. Correlations between circulating adipocytokine levels and those of multiple well-established markers of insulin resistance including homeostatic model assessment of insulin resistance (HOMA-IR), homeostatic model assessment of β-cell function (HOMA-β) and quantitative insulin sensitivity check index were also established.

Results

CRP, IL-8, MCP-1, and TNF-α levels were higher in obese than non-obese individuals; the CRP and IL-8 differences were statistically significant. CRP correlated significantly with markers of insulin resistance (fasting plasma insulin, HOMA-IR, and QUICKI), and adiponectin correlated with HOMA-β in obese individuals. We divided the group of obese individuals on the basis of HOMA-IR levels into insulin-resistant (IR; HOMA-IR ≥2.5) and insulin-sensitive (IS; HOMA-IR <2.5) groups; and 43 out of 62 participants were IR despite comparable BMIs. An overall proinflammatory profile was compared between IR and IS obese, though the values were higher in IR obese but the difference was not significant.

Conclusion

Obesity is associated with a general inflammatory milieu and a crosstalk between adipocytokines and insulin resistance is complex as well as multifactorial.

INTRODUCTION

The rising incidence and prevalence of overweight and obesity are posing a major challenge to chronic disease prevention and public health across countries.¹ Despite significant advances in our understanding of the underlying mechanisms of weight gain and associated comorbidities like insulin resistance and cardiovascular risk, the prevalence of obesity and metabolic syndrome is escalating at an unprecedented rate.² Worldwide, obesity has nearly tripled since 1975. Presently, the majority of the world’s population lives in countries in which obesity is responsible for more fatalities than malnutrition and underweight status.³ Overweight conditions and obesity, defined as a body mass index (BMI) of ≥30 kg/m², had highest prevalence (42.4%) in the United States in 2017–2018.4 India is also witnessing a mounting prevalence of these conditions as more than 135 million individuals are affected by obesity,⁵ particularly abdominal obesity. Additionally, Asians often display higher fat percentages at similar BMIs when compared to their counterparts from the western world.⁶

The genesis of the adipose tissue dysfunction in obesity is complex. A review of the scientific literature suggests that in response to changes in the nutritional cues in the setting of varied genetic and environmental factors, the adipose tissue undergoes dynamic remodelling.² This is evident in terms of quantitative and qualitative alterations in the population of adipocytes, thereby placing these cells under an immense metabolic and oxidative stress.² This remodeling in adipose tissue is closely related to the changes in tissue functions like alterations in adipokine secretion, local hypoxia and adipocyte death, inflammation and worsening of metabolic health.⁷ Active secretion of a variety of adipocytokines, which modulate metabolic homeostasis, has been observed in obesity.⁸ These adipocytokines are measurable in the circulation and have been known to impart a systemic low-grade chronic inflammation profile to obese and overweight individuals. Adoption of a more comprehensive term, “meta-inflammation” or metabolically triggered inflammation, has been suggested which points towards the genesis of the inflammation.⁹ This inflammation is considered to be the major risk factor for the development of obesity-related comorbidities like insulin resistance and cardiovascular diseases.¹⁰

Many population-based studies performed to date have broadly stressed that an imbalance of pro- and anti-inflammatory adipocytokines impacts the metabolic outcomes in obesity in terms of associated comorbidities.⁷ However, the studies have not reported the levels of individual adipocytokines in obesity and metabolic syndrome consistently. This has made interpretation difficult, and the underlying reasons seem more complex than expected.^11,12 An understanding of the inflammatory and metabolic crosstalk of the cytokines in the trajectory of obesity needs further exploration.

Therefore, we designed our study to compare the levels of adipocytokines in morbidly obese individuals undergoing bariatric surgery with those of non-obese individuals. We used a multiplexed array of the most potent adipokines known to affect insulin sensitivity, i.e., adiponectin and resistin, and other serum inflammatory cytokines like C-reactive protein (CRP), interleukins (IL)-6 and -8, monocyte chemoattractant protein (MCP)-1 and tumor necrosis factor (TNF)-α. We chose the Luminex multiplex immunoassay in combination with drop array technology for our study considering its advantages over conventional enzyme-linked immunosorbent assay (ELISA).^13,14

Modulation of inflammatory status in obesity seems fundamental to its complex relationship with insulin response. Thus, by using the calculated parameters of insulin sensitivity including homeostatic model assessment of insulin resistance (HOMA-IR),¹⁵ the marker of homeostatic model assessment of β-cell function (HOMA-β)16 and the quantitative insulin sensitivity check index (QUICKI),¹⁷ we aimed to establish an association between circulating adipocytokines and insulin sensitivity in obesity.

METHODS

Study design and participant recruitment

We designed an exploratory cross-sectional study in which participants were recruited through the elective surgery schedule of the Department of Surgical Disciplines (Laparoscopic and Bariatric Surgery), All India Institute of Medical Sciences (AIIMS), New Delhi. We included 62 adult individuals with obesity (BMI ≥30 kg/m²) who were admitted for undergoing bariatric surgery. The control group included 32 non-obese (BMI <25kg/m²) adult individuals undergoing other surgeries with associated disease conditions. These included laparoscopic cholecystectomy for cholelithiasis and incisional hernia repair. Individuals younger than 18 years and older than 60 years were excluded. We also excluded candidates diagnosed with diabetes mellitus and concomitant acute or chronic disorders of the immune system. Written informed consent was obtained from all participants, and the study was approved by the AIIMS Ethical Committee (A-536; IEC no. 473/01.09.2017, RP-60/2017).

Anthropometric measurements

All subjects underwent comprehensive medical evaluation including history and physical examination prior to surgery. The age, sex, height (cm), weight (kg), and waist circumference (cm) were recorded, and BMI was calculated as weight (kg)/height (m²) for all participants.

Sample collection and processing

Preoperative blood samples were obtained by venipuncture after overnight fasting in participants in both study groups. Appropriate vacutainer tubes were used for the measurement of plasma glucose, insulin and serum cytokines. After centrifugation at 3,000 rpm for 15 minutes, plasma and serum were separated. Plasma was used for measuring glucose and insulin, and serum was transferred and stored in sterile Eppendorf tubes at –80°C for subsequent analysis.

Plasma glucose and insulin assays and serum glycosylated hemoglobin measurement

Plasma glucose was estimated using Randox GOD-PAP glucose estimation kit (Randox Laboratories, Crumlin, UK) according to manufacturer’s protocol.¹⁸ Insulin was estimated using chemiluminescence-based immunoassay in a Liaison autoanalyzer (Diasorin, Saluggia, Italy). Serum glycosylated hemoglobin was determined using high performance liquid chromatography.¹⁸

To test for associations between the inflammation profile and insulin sensitivity status of the individuals in our study, we used the well-established markers of insulin resistance HOMA-IR, HOMA-β, and QUICKI as useful indices for insulin sensitivity. These indices are calculated parameters and use the value of fasting plasma glucose and insulin for their determination. HOMA-IR was calculated by using the formula (HOMA-IR)=insulin (µU/mL)×[glucose (mmol/L)/22.5].¹⁵ Low HOMA-IR (<2.5) indicated insulin sensitivity, and high HOMA-IR values (≥2.5) indicated insulin resistance.^18,19 We used this cutoff level of HOMA-IR to further subgroup obese individuals as insulin-sensitive (IS) and insulin-resistant (IR) and compared the inflammatory status between the two. HOMA-β was calculated as 20×insulin (µU/mL)/[glucose (mmol/L)–3.5].¹⁶ QUICKI was calculated as 1/[log(I0)+log(G0)], where I0 is fasting insulin (µU/mL) and G0 is fasting glucose (mg/dL).¹⁷ QUICKI is one of the most accurate and useful indices for determining insulin sensitivity; a value of below 0.34 suggests insulin resistance and below 0.30 supports the diagnosis of diabetes mellitus.¹⁷

Serum adipocytokine assay

Serum samples were subjected to quantitative analysis of the circulating adipocytokines. A panel was made for measuring serum adipokines and inflammatory cytokines. This panel consisted of adiponectin, resistin, CRP, IL-8, MCP-1, IL-6, and TNF-α. These were assayed using R&amp;D Systems Luminex Multiplex Immunoassay (Minneapolis, MN, USA) as per the manufacturer’s protocol in combination with Drop array technology and read in a Bio-Plex 200 reader (Bio-Rad, Hercules, CA, USA). This system makes use of xMAP magnetic bead microspheres which act as both the identifier and the solid surface to build the assay. In this system, multiplex immunoassays consisting of analyte-specific capture antibodies are conjugated to xMAP beads. A combination of analyzer and software for data acquisition and analysis allows higher throughput and accuracy than conventional immunoassays with sample volumes as low as 25 µL. Multiplexing also minimizes reaction to reaction inconsistencies as might be seen with a separate ELISA for each cytokine.^13,14

Statistical methods

To describe patient demographics and anthropometric characteristics and biochemical parameters, the data was summarized and analyzed using Prism version 8.0.1 (GraphPad, San Diego, CA, USA). Data were tested for normality using Shapiro-Wilk/Kolmogorov-Smirnov test. Continuous data was expressed as mean±standard error of the mean. Student t-test was used to compare the normally-distributed variables between study groups, and Mann-Whitney U-test was performed to compare the non-normally-distributed variables. A P-value of less than 0.05 was considered to be statistically significant. Correlation between anthropometric characteristics and markers of insulin resistance with the levels of circulating adipocytokines was calculated using Spearman’s correlation coefficient. The P-values <0.05 were considered to be statistically significant.

RESULTS

Clinical characteristics of participants

Average BMI of the individuals in the obese study group was 45.76±0.81 kg/m² and in the non-obese control group 24.19±0.80 kg/m². Average age of individuals in the obese and non-obese group was 39.55±1.38 and 38.91±2.63 years, respectively (Table 1). The sex distribution was 53 females and 9 males in the obese group and 20 females and 12 males in the non-obese group.

Table 1.

Anthropometric characteristics, biochemical parameters and serum adipocytokines in obese and non-obese groups

Variable	Obese (n = 62)	Non-obese (n = 32)	P
Age (yr)*	39.55 ± 1.38	38.91 ± 2.63	NS
Anthropometric characteristics*
Body weight (kg)	115.8 ± 3.06	61.14 ± 2.01	< 0.001
BMI (kg/m2)	45.76 ± 0.81	24.19 ± 0.70	< 0.001
Clinical parameter*
Fasting plasma glucose (mg/dL)	107.1 ± 3.02	95.26 ± 2.81	0.006
Fasting plasma insulin (μU/L)	21.86 ± 2.36	10.73 ± 1.67	< 0.001
HOMA-IR	5.38 ± 0.65	2.17 ± 0.24	< 0.001
HOMA-β	242.4 ± 33.45	115.3 ± 18.56	0.001
QUICKI	0.31 ± 0.00	0.48 ± 0.12	0.035
Serum adipocytokine†
CRP (μg/mL)	52.56 ± 15.55	34.05 ± 19.09	0.006
Adiponectin (μg/mL)	39.88 ± 4.17	44.09 ± 7.33	0.593
IL-8 (pg/mL)	25.80 ± 38.45	36.39 ± 6.45	0.025
MCP-1 (ng/mL)	1.090 ± 0.17	0.75 ± 0.05	0.063
Resistin (ng/mL)	39.70 ± 5.98	36.53 ± 5.86	0.707
IL-6 (pg/mL)	26.88 ± 6.99	37.72 ± 32.72	< 0.001
TNF-α (pg/mL)	38.20 ± 7.88	23.05 ± 3.63	0.084

Values are presented as mean± standard error of the mean.

The P-values obtained by using ^*Student t-test (unpaired) for parametric; ^†Mann-Whitney U-test for non-parametric variables between two groups (P<0.05 is considered significant).

NS, not significant; BMI, body mass index; HOMA-IR, homeostatic model assessment of insulin resistance; HOMA-β, homeostatic model assessment of β-cell function; QUICKI, quantitative insulin sensitivity check index; CRP, C-reactive protein; IL, interleukin; MCP, monocyte chemoattractant protein; TNF, tumor necrosis factor.

Biochemical parameters and markers of insulin resistance

All individuals in our study were non-diabetic, but we found significantly higher fasting plasma glucose and insulin levels in obese as compared to non-obese participants. All markers of insulin resistance, i.e., HOMA-IR, HOMA-β, and QUICKI, showed that the participants in the obese group were significantly more insulin-resistant than those in the non-obese group (Table 1). Since one of the objectives of our study was to assess how inflammatory status differed with the insulin response in obese individuals, we divided the participants in this group into IS and IR obese subgroups based on the HOMA-IR cutoff as previously mentioned in the Methods section.^18,19 Out of 62 obese individuals in our study, 43 were IR and 19 were IS (Table 2). The difference in their BMI was not significant; however, all three insulin sensitivity markers, i.e., HOMA-IR, HOMA-β, and QUICKI, showed significant differences between IR and IS obese groups (Table 2).

Table 2.

Anthropometric characteristics, biochemical parameters and serum adipocytokines in IR obese and IS obese subgroups

Variable	IR obese (n = 43)	IS obese (n = 19)	P
Age (yr)*	39.07 ± 1.66	40.63 ± 2.53	NS
Anthropometric characteristics*
Body weight (kg)	116.40 ± 3.92	114.30 ± 4.44	0.762
BMI (kg/m2)	46.20 ± 1.01	44.66 ± 1.34	0.399
Clinical parameter*
Fasting serum glucose (mg/dL)	107.40 ± 3.76	106.20 ± 5.02	0.862
Fasting serum insulin (μU/L)	25.74 ± 2.83	9.949 ± 1.76	0.003†
HOMA-IR	6.53 ± 0.74	1.35 ± 0.15	< 0.001
HOMA-β	284.40 ± 40.37	95.40 ± 21.72	0.017
QUICKI	0.30 ± 0.003	0.35 ± 0.01	< 0.001
HbA1c	6.07 ± 0.09	5.83 ± 0.26	0.297
Serum adipocytokines†
CRP (μg/mL)	60.79 ± 20.04	33.90 ± 22.92	0.430
Adiponectin (μg/mL)	52.16 ± 6.61	72.17 ± 12.27	0.124
IL-8 (pg/mL)	152.40 ± 53.40	65.82 ± 31.69	0.179
MCP-1 (ng/mL)	1.05 ± 0.16	1.18 ± 0.43	0.177
Resistin (ng/mL)	38.37 ± 6.60	42.71 ± 12.8	0.741
IL-6 (pg/mL)	32.21 ± 9.70	14.82 ± 5.64	0.091
TNF-α (pg/mL)	35.45 ± 6.49	44.42 ± 21.46	0.510

Values are presented as mean± standard error of the mean.

The P-values obtained by using ^*Student t-test (unpaired) for parametric; ^†Mann-Whitney U-test for non-parametric variables between two groups (P<0.05 is considered significant).

IR, insulin-resistant; IS, insulin-sensitive; NS, not significant; BMI, body mass index; HOMA-IR, homeostatic model assessment of insulin resistance; HOMA-β, homeostatic model assessment of β-cell function; QUICKI, quantitative insulin sensitivity check index; HbA1c, glycosylated hemoglobin; CRP, C-reactive protein; IL, interleukin; MCP, monocyte chemoattractant protein; TNF, tumor necrosis factor.

Serum adipocytokine assay

Out of the five inflammatory cytokines that we assayed in our study, we found a higher level of CRP, IL-8, MCP-1, and TNF-α in obese as compared to non-obese individuals; the difference, however, was significant only for CRP and IL-8 (Table 1). Surprisingly, IL-6 was found to be significantly lower in obese than in non-obese individuals, contrary to the data generally reported in most of the literature.^2,20 Considering the two adipokines that were part of the same array, resistin was higher and adiponectin was lower in obese participants than in non-obese ones; the differences, however, were not statistically significant (Table 1).

Our next aim was to assess how the inflammatory status with obesity changed with whole body insulin resistance. We found that neither of the adipocytokines showed a significant difference between IR and IS obese groups. Overall, though, a presence of increased proinflammatory profile was evident for the IR obese participants due to their higher levels of CRP, IL-8, and IL-6 than the IS obese group participants. This finding suggests that despite being non-diabetic clinically, 43 out of 62 of the IR obese individuals presented with a proinflammatory profile (Table 2). Those in this group are at higher risk of developing metabolic and cardiovascular complications.

Correlation of inflammatory profile of obese with anthropometric characteristics and markers of insulin resistance

The analysis of correlations was done using Spearman’s correlation coefficient because many variables were not normally-distributed in different groups. Out of the cytokines studied, IL-8 correlated significantly with BMI of obese (r=0.215, P=0.04) and IS obese individuals (r=0.483, P=0.02). Only CRP correlated significantly with fasting insulin, HOMA-IR and QUICKI in obese and IR obese individuals. Also, CRP was the only cytokine that correlated significantly with HOMA-IR in obese (r=0.382, P=0.002), IR obese (r=0.380, P=0.006) and IS obese (r=0.678, P=0.009) groups and subgroups respectively. There was also a significant positive correlation between TNF-α and fasting plasma glucose in the obese group (r=0.262, P=0.02) and IR obese groups (r=0.312, P=0.02) (Tables 3 and 4). Therefore the overall proinflammatory milieu in obesity correlated significantly with insulin resistance.

Table 3.

Correlations among circulating adipocytokines, anthropometric characteristics and markers of insulin resistance in obese individuals undergoing bariatric surgery

Variable	Weight	BMI	Fasting plasma glucose	Fasting plasma insulin	QUICKI	HOMA-IR	HOMA-β	HbA1c
CRP	–0.099	0.030	–0.035	0.289*	–0.239*	0.382†	0.150	0.038
Adiponectin	–0.145	–0.051	0.305*	0.034	–0.204	0.251*	–0.134	–0.042
IL-8	0.132	0.215*	0.078	0.008	–0.018	0.203	–0.001	0.122
MCP-1	–0.010	0.104	–0.099	0.043	0.005	–0.018	0.053	–0.049
Resistin	0.004	–0.175	–0.137	–0.012	0.115	–0.106	0.082	–0.177
IL-6	0.016	0.069	–0.205	0.104	–0.061	0.081	0.153	0.186
TNF-α	–0.107	0.015	0.262*	0.118	–0.136	0.185	–0.082	0.054

Spearman correlation coefficient represented as r-value: ^*P<0.05; ^†P<0.01.

BMI, body mass index; QUICKI, quantitative insulin sensitivity check index; HOMA-IR, homeostatic model assessment of insulin resistance; HOMA-β, homeostatic model assessment of β-cell function; HbA1c, glycosylated hemoglobin; CRP, C-reactive protein; IL, interleukin; MCP, monocyte chemoattractant protein; TNF, tumor necrosis factor.

Table 4.

Correlations among circulating adipocytokines with anthropometric characteristics and markers of insulin resistance in IR and IS obese individuals undergoing bariatric surgery

Variable	Weight	BMI	Fasting plasma glucose	Fasting plasma insulin	QUICKI	HOMA-IR	HOMA-β	HbA1c
CRP
IR	–0.010	0.013	0.081	0.282*	–0.306*	0.380†	0.104	0.016
IS	–0.308	0.046	–0.365	0.039	0.281	0.678†	–0.034	0.025
Adiponectin
IR	–0.158	-	0.503†	–0.063	–0.171	0.160	–0.323*	0.004
IS	–0.134	–0.301	–0.344	0.164	–0.277	0.545*	0.433	0.088
IL-8
IR	0.018	0.040	0.113	–0.145	0.084	-	–0.101	0.206
IS	0.394	0.483*	–0.049	–0.048	0.288	0.165	0.010	–0.177
MCP-1
IR	–0.107	0.003	–0.161	–0.069	0.120	–0.112	0.053	–0.085
IS	0.188	0.369	0.157	–0.090	0.327	–0.006	–0.174	-0.060
Resistin
IR	–0.204	–0.419†	–0.176	0.032	0.103	–0.024	0.146	–0.146
IS	0.532*	0.564*	0.085	0.129	0.145	–0.062	0.111	–0.155
IL-6
IR	0.031	-	–0.156	–0.012	–0.007	–0.025	0.125	0.204
IS	–0.075	0.087	–0.436*	–0.020	0.410	–0.504*	0.192	0.018
TNF-α
IR	–0.157	–0.009	0.312*	–0.130	0.053	0.058	–0.206	0.093
IS	0.025	–0.079	0.091	0.453	–0.233	0.229	0.129	–0.074

Spearman correlation coefficient represented as r-value: ^*P<0.05; ^†P<0.01.

IR, insulin-resistant; IS, insulin-sensitive; BMI, body mass index; QUICKI, quantitative insulin sensitivity check index; HOMA-IR, homeostatic model assessment of insulin resistance; HOMA-β, homeostatic model assessment of β-cell function; HbA1c, glycosylated hemoglobin; CRP, C-reactive protein; IL, interleukin; MCP, monocyte chemoattractant protein; TNF, tumor necrosis factor

With respect to the two adipokines studied in our participants, adiponectin correlated significantly with pancreatic insulin secretion capacity demonstrated by its negative correlation with HOMA-β in the IR obese subgroup (r=–0.323, P=0.01) (Table 4). Resistin showed a significant positive correlation with weight (r=0.532, P=0.014) and BMI (r=0.564, P=0.01) in the IS obese group but an unexpected negative correlation with BMI (r=–0.419, P=0.002) in the IR obese (Tables 3 and 4).

DISCUSSION

In our study, we sought to evaluate the distinct pattern of an array of circulating adipocytokines in obese as compared to non-obese individuals. We designed a cross-sectional study in which we also intended to establish a correlation between the inflammatory profile of the obese subjects and insulin response. All participants in our study were non-diabetic, but the markers of insulin resistance (HOMA-IR, HOMA-β, and QUICKI) significantly showed that obese individuals were more insulin-resistant than non-obese ones (Table 2). We measured two important adipokines (adiponectin and resistin) in all participants of our study and considered their opposite roles in mediating the insulin response; adiponectin has been implicated in increasing insulin sensitivity, and resistin has been linked to obesity with insulin resistance.^21,22 We found that obese individuals had lower adiponectin levels than non-obese ones. The preponderance of scientific evidence suggests that adiponectin level has an inverse association with adiposity, and higher levels of adiponectin have been observed in lean individuals.²³ In addition, many studies have shown that adiponectin level is lower in the insulin-resistant state and increases with improvement in the insulin sensitivity associated with the weight loss.²¹ We also found a lower level of adiponectin in IR obese individuals as compared to IS obese ones in our study, but the difference was not statistically significant (Table 2). However, adiponectin correlated significantly with HOMA-β in IR obese individuals, but its correlation with other markers of insulin resistance was inconclusive across the groups in our study (Tables 3 and 4). Similar observations were made in other studies, and an “adiponectin paradox” in which epidemiological data demonstrated that increased adiponectin levels correlated with higher cardiovascular or all-cause mortality was reported.²⁴ The underlying mechanism behind this paradox is still unclear, but one hypothesis is that there may be a compensatory elevation of adiponectin in patients with metabolic abnormalities. This compensatory elevation may be associated with higher future mortality.²⁴

The other adipokine analyzed in our study was resistin which has an opposite effect as that of adiponectin in terms of insulin sensitivity. Experimental studies have shown that resistin increases hepatic gluconeogenesis and insulin resistance.²⁵ We did not find a significant difference in the levels of this adipokine in either obese vs. non-obese or IR versus IS obese groups. Population-based studies have shown that the association between resistin levels and insulin sensitivity has been inconsistent.²⁴ Furthermore, resistin’s role in insulin resistance and glucose metabolism has been inconclusive in human studies.²⁶ The only significant correlations that we found for resistin were with weight and BMI, and those associations were only in the IS obese group. This finding is in agreement with several studies reporting that increased plasma resistin levels were associated with increased BMI.²⁶

Apart from their opposite roles in mitigating insulin response, both adiponectin and resistin also have opposite effects on inflammation profiles. Adiponectin has anti-inflammatory and anti-atherogenic effects²¹ by suppressing proinflammatory cytokines such as TNF-α and by inducing other anti-inflammatory factors.²¹ However, resistin has been shown to exert proinflammatory effects by increasing the production of IL-6 and TNF-α.²⁷

We included inflammatory cytokines in the array of adipocytokines in our study and found that CRP, IL-8, MCP-1, and TNF-α were higher in obese than non-obese participants. CRP and IL-8 level were significantly different suggesting that obesity is a chronic inflammatory state. However, IL-6, despite being proinflammatory, was found to be significantly lower in obese than non-obese individuals in our study. There are several potential explanations for this ostensibly contradictory data for the role of IL-6 in glucose metabolism and insulin action.¹⁰ Researchers disagreed on whether IL-6 correlated with BMI, insulin levels or HOMA-IR.²⁸ Many experimental studies have determined that there exists a fine-tuned balance between anti-and proinflammatory actions of IL-6.²⁹ Additionally, IL-6 is produced by different organs. This may contribute to the layers of complexity associated with its actions on metabolic regulation.¹⁰ Owing to such inconsistencies, a pleiotropic functional profile has been suggested for IL-6.³⁰

Another IL measured in our study was IL-8, and we found significantly higher levels of IL-8 in obese than in non-obese individuals. This also applied to the IR versus IS obese subgroups. IL-8 also showed a significant positive correlation with BMI in both obese and IR obese groups. Some of the recent studies have clearly demonstrated that higher IL-8 levels are associated with obesity.¹³ IL-8 levels are also closely correlated with obesity-related parameters like BMI, waist circumference, CRP, IL-6 levels and HDL cholesterol levels.³¹

One of the important and most abundant immune cells infiltrating the adipose tissue in obesity are macrophages,¹⁰ and MCP-1 is an important chemoattractant chemokine produced by adipose tissue mediating macrophage recruitment.² We found increased circulating levels of MCP-1 in obese individuals as compared to non-obese ones, but the difference was not statistically significant. Dahlman et al.³² found that expression of MCP-1 is significantly higher in obese as compared to lean individuals, but circulating levels showed no changes among their study groups suggesting that obesity produced only local changes of MCP-1 in adipose tissue.

There is putative evidence that an imbalance of the pro-and anti-inflammatory adipokines secreted by adipose tissue is one of the major contributors to metabolic dysfunction. However, there are inconsistencies in the outcomes of a number of human studies.^11,33 We did not find a significant difference but observed a notable trend in which classical proinflammatory cytokines like CRP, IL-8, and IL-6 were found to be higher and MCP-1 and TNF-α were paradoxically lower in IR as compared to IS obese study participants.

TNF-α has been the most consistently studied inflammatory cytokine in obesity ever since its identification in the adipose tissue of obese mice. This marked the start of the metabolic inflammation concept.^10,34 TNF-α levels did not show any significant difference between the groups, but correlated significantly with fasting plasma glucose in obese and IR obese individuals in our study. Strong associations have been shown in several human studies between circulating TNF-α and insulin resistance³⁵ or other metabolic complications.³⁶ However, attempts to block TNF-α functions have never produced consistent metabolic outcomes.¹⁰ Adipose tissue does not significantly contribute to circulating TNF-α levels.³⁷ In fact, the soluble forms of the receptor of TNF-α (TNFR2) is secreted from adipose tissue; this may modulate the actions of TNF-α and warrants detailed exploration.³⁷

Functional pleiotropy and redundancy are characteristic features of cytokines. As cytokines are produced by many different cell types, cytokines often show overlapping activities in regulating proliferation or differentiation; these regulatory activities depend on the type and developmental state of the target cells involved.³⁰ Collectively, the signaling of TNF-α and IL-6 in obesity-associated inflammation is complex, especially in developing concomitant insulin resistance, owing to their temporal and cell-type specific functions.²⁹ Considering that resistin increases IL-6 and TNF-α secretion along with that of other adhesion molecules,²⁴ we may assume that inconsistencies in resistin levels may also be responsible for unexpected results and correlations for these two cytokines.

Fundamental differences in mechanisms that regulate inflammatory cytokines (TNF-α, IL-6, and IL-8) and acute phase proteins like CRP may explain the inconsistency of relationships between these markers and metabolic phenotypes in morbidly obese individuals.^20,28 Equally possible is that technical differences in sample collection/handling may be responsible for varying outcomes.

Of all the cytokines we studied, only CRP correlated significantly with most of the markers of insulin resistance (HOMA-IR, QUICKI, and fasting insulin) and had significantly higher levels in obese and IR obese individuals. CRP is the most often measured inflammatory marker and has consistently correlated with weight, BMI and waist to hip ratio as well as insulin resistance.^2,20 CRP has become the marker of metainflammation.² This may be attributed to its crosstalk with other cytokines like TNF-α and IL-6.^2,20 Despite the extensive characterization of the mechanisms by which inflammatory cytokines promote insulin resistance and adipose tissue dysfunctions, targeted immunotherapies have not yielded any positive outcome.¹¹ Driving forces behind changes in inflammatory markers are multifactorial. A chronology of the triggers for each of the cytokines that can better establish the trajectory of presentation and inter- and intra-individual variations is lacking.

Based on the available literature and the findings of our study, obesity is associated with a general inflammatory response that cannot be simply attributed to mere imbalance between pro- and anti-inflammatory status. A comprehensive understanding of the crosstalk between inflammation and insulin response with the quantum of effect of individual cytokines is necessary to provide a better insight into the functions of adipose tissue as an essential metabolic regulator.