Cytokines and leptin correlation in patients with polycystic ovary syndrome: Biochemical evaluation in south Indian population

Abstract

Background:  In polycystic ovary syndrome (PCOS), the leptin (OB protein) is related to reproductive function and inflammatory response. Leptin and cytokines have been thought to be putative local regulators in PCOS.

Methods:  To examine the relationship between serum leptin and serum interleukin‐6 (IL‐6), interleukin‐8 (IL‐8) and tumor necrosis factor‐α (TNF‐α) levels in underweight, overweight, obese and morbidly obese PCOS and non‐PCOS subjects compared with normal weight, regularly menstruating women.

Results:  Leptin levels are highly correlated with TNF‐α, IL‐6 and IL‐8. There is a significant dependent increase with the degree of obesity, but in underweight PCOS subjects, leptin levels are elevated irrespective of the body mass index.

Conclusion:  The present study showed that leptin levels were elevated in underweight and morbidly obese PCOS subjects. This could be the result of impaired expression of leptin in PCOS, leading to leptin resistance. As a result of this regulation, TNF‐α, IL‐6 and IL‐8 were also elevated in morbidly obese and underweight PCOS subjects. In obese subjects, where there was an increase in adipose mass, increased levels of leptin were observed and this was attributed to the inflammatory properties while increasing the adipose mass. Serum IL‐6 and IL‐8 circulate at high levels and are more important systemically. They are, perhaps, the hormonal factors that induce leptin and insulin resistance in underweight PCOS subjects. Therefore, leptin and inflammatory markers were acting at paracrine and endocrine levels in PCOS subjects. (Reprod Med Biol 2005; 4: 247–254)

INTRODUCTION

OBESITY IS AMONG the most common health problems in industrialized societies. Obesity, which is the result of an imbalance between caloric intake and energy expenditure, is highly correlated with leptin and insulin resistance in PCOS subjects. It has also been suggested that leptin serves as a permissive signal to reproductive functions. ¹ In addition to these effects of leptin on the reproductive axis, there are several additional reasons to assess the role of leptin in PCOS. In PCOS subjects, serum leptin levels correlate well with bodyweight, body mass index (BMI), percentage of body fat, serum levels of TNF‐α, IL‐6 and IL‐8. Obese subjects generally have higher levels of serum leptin, suggesting that obese people are not defective in leptin expression but, rather, manifest leptin resistance. ² Several studies have noted increased ob mRNA expression in obese people ³ and the elevated serum leptin levels are consistently directly related to adipose tissue ob gene expression. Another important product of adipocytes is TNF‐α, which is elevated in the adipose tissue of obese rodents and humans, ⁴ and might play an important role in obesity‐related insulin resistance. ⁵ Although adipocyte TNF‐α concentration is related to obesity, there is much interindividual variation in humans, suggesting that other factors control TNF‐α expression. Both insulin resistance and TNF‐α overexpression in adipose tissue and skeletal muscle are important features of human obesity, related to each other, as TNF‐α induces insulin resistance by acting through the autocrine–paracrine pathway. ⁶ It is well known that insulin resistance plays a role in the pathogenesis of obesity‐related complications. Cytokine IL‐8 is produced mainly in macrophages and monocytes and plays a role in modulating the inflammatory response. Oxidized low‐density lipoprotein (LDL) particles are able to stimulate production and secretion of IL‐8 by macrophages from human atherosclerotic plaques and high levels of IL‐8 is in macrophage‐derived human foam cells. ⁷ IL‐8 is a potent chemoattractant and is responsible for the recruitment of neutrophils and T lymphocytes into the subendothelial space. It also induces adhesion of monocytes to endothelium and migration of vascular smooth muscle cells. ⁸ Another adipocyte secretory product that might be involved in insulin resistance is interleukin (IL‐6, IL‐8), which is a cytokine secreted by many cells, including adipocyte and adipose stromal cells. Like TNF‐α, IL‐6 and IL‐8 inhibit the expression of LPL but unlike TNF‐α, IL‐6 and IL‐8 do not stimulate lipolysis. ⁹ IL‐6 and IL‐8 secretion is increased in the adipocytes of obese subjects and might be important as either a circulating hormone or as a local regulator of insulin action. Although many studies have examined the role of TNF‐α in insulin resistance, relatively few of these studies have used human subjects, and none has examined cytokine expression in detail along with the measurement of insulin resistance in obese subjects. ¹⁰ In the present study, we examined the serum levels of leptin, TNF‐α, IL‐6 and IL‐8 in underweight, overweight, obese and morbidly obese women with PCOS subjects and compared these with normal weight subjects. We found that circulating leptin, TNF‐α, IL‐6 and IL‐8 levels, which ultimately lead to defects in leptin action in female infertility, are highly correlated with obese PCOS subjects and underweight PCOS subjects than in regularly menstruating, normal weight subjects.

MATERIALS AND METHODS

Subjects

THE STUDY POPULATION consisted of 20 underweight non‐PCOS subjects and 20 underweight PCOS subjects (BMI 20.0 kg/m²), 25 overweight non‐PCOS subjects and 25 overweight PCOS subjects (BMI 32.2 kg/m²), 25 obese non‐PCOS subjects and 25 obese PCOS subjects (BMI 38.0 kg/m²), 40 morbidly obese non‐PCOS subjects and 40 morbidly obese PCOS subjects (BMI 43.0 kg/m²), and 100 regularly menstruating, normal weight control women (BMI 27.0 −27.8 kg/m²). Women with PCOS who were less than 45 years of age (18–44 years) with polycystic ovarian morphology were confirmed by visual inspection of the ovaries at laparotomy, laparoscopy or by ultrasound examination.

There was no evidence of hyperprolactinemia, Cushing's syndrome, congenital or non‐classical adrenal hypoplasia and hormone secreting tumors. Control women were healthy, less than 45 years of age (18–44 years), with a regular menstrual cycle and no evidence of hyperandrogenism, polycystic ovaries, endometriosis or abnormal uterine bleeding.

Neither PCOS nor control women were under medication within 60 days prior to the date of blood collection from them. After obtaining written consent, fasting blood was drawn at the Institute of Obstetrics and Gynecology (Madras Medical College, Chennai) and University centers (Institute of Basic Medical Services). The subjects were allowed to fast for 12 h and then 2 mL of blood was collected from them. The blood was allowed to clot and it was retracted and separated by centrifugation at 2000 g for 15 min. The serum was separated carefully and stored at −80°C until analysis.

Radioimmunoassay method for serum leptin concentration

The circulating leptin concentration was measured using a kit Human Leptin IRMA Kit ¹¹ (Diagnostic Systems Laboratories, Webster, TX, USA).

Serum TNF‐α concentration

TNF‐α was measured by the TiterZyme enzyme immunometric assay (Assay Designs, Ann Arbor, MI, USA) with the minimum detectable concentration of 1.9 pg/mL and with the intra‐assay coefficient of variation less than 8.3%. ¹²

Serum human IL‐6 concentration

Serum human IL‐6 concentration was assessed with the sandwich immunoassay kit for the detection of ‘free’ human IL‐6 (CHEMICON International, Muenchen, Germany). ¹³

Serum human IL‐8 concentration

Serum human IL‐8 concentration was measured by the sandwich immunoassay kit for the detection of ‘free’ human IL‐8 (PromoCell GmbH, Heidelberg, Germany) with the standard curve range of 0.40–24.0 pg/mL and the detection limit of the method less than 97 pg/mL. ¹⁴ The intra‐assay and interassay coefficient of variation were less than 7.3% and 8.9%, respectively.

Serum biochemical analysis

Serum insulin was measured by radioimmunoassay ¹⁵ (Board of Radiation and Isotope Technology BARC Vashi complex, Navi Mumbai, India). Serum glucose was estimated by the glucose oxidase–peroxidase method, ¹⁶ serum cholesterol was measured by the cholesterol oxidase–peroxidase method ¹⁷ serum high density lipopoetein (HDL)‐cholesterol was measured by the direct method ¹⁸ and serum triglycerides was measured by the glycerol‐3‐phosphate oxidase–peroxidase–N‐ethyl‐methylanilin propan‐sulphonate sodic method ¹⁹ using autoanalyzer (BAYER RA 50; Bayer Company India, Guindy, Chennai, India).

Statistical analysis

Univariate linear regression was used to examine the relationships between BMI, percentage of body fat and leptin concentrations, simultaneously controlling for age. Student's t‐test was carried out to analyze the differences between groups at specific time‐points; the linear regression equation and correlation coefficient were determined by least square analyses. The results were presented as mean ± SD. The analysis was two‐tailed and carried out using spss 7.5.

RESULTS

Relationship between leptin and TNF‐α, IL‐6 and IL 8 in obesity related polycystic ovary syndrome and non‐polycystic ovary syndrome subjects

IN THE PRESENT study, 320 women including normal weight control subjects, underweight, overweight, obese and morbidly obese PCOS and non‐PCOS subjects and their serum levels of leptin, TNF‐α, IL‐6 and IL‐8 concentrations were analyzed and they are shown in1, 2. The observation of the lowest values for leptin, TNF‐α, IL‐6 and IL‐8 were 2.0 ng/mL, 134.0 pg/mL, 17.0 pg/mL, 16.5 pg/mL, respectively. The highest values were 43.0 ng/mL, 1324.7 pg/mL, 1542.7 pg/mL, 1156.4 pg/mL with a 22‐fold, 10‐fold, 90‐fold and 70‐fold difference, respectively.1, 2 represent the levels of leptin, BMI, TNF‐α, IL‐6 and IL‐8 in underweight, overweight, obese, morbidly obese, non‐PCOS and normal weight control subjects. The measured serum leptin and serum TNF‐α, IL‐6 and IL‐8 levels were two to three times higher in underweight, overweight, obese and morbidly obese PCOS women than in normal weight control and underweight non‐PCOS subjects. A significant difference was observed in leptin, TNF‐α, IL‐6, IL‐8, glucose, insulin, triglycerides, cholesterol and HDL‐cholesterol between PCOS and non‐PCOS of underweight and morbidly obese subjects compared with normal weight subjects. To better define the effects of obesity on TNF‐α, we measured TNF‐α from each PCOS subject, along with IL‐6 and IL‐8. Subjects were divided into five BMI groups representing underweight (BMI = 20.0 kg/m², PCOS = 20, non‐PCOS = 20), normal weight (BMI = 27.8 kg/m², n = 100), overweight (BMI = 32.2 kg/m², PCOS = 25, non‐PCOS = 25), obese (BMI = 38.0 kg/m², PCOS = 25, non‐PCOS = 25), morbidly obese (BMI = 43.0 kg/m², PCOS = 40, non‐PCOS = 40). Figure 2 shows TNF‐α levels from subjects with increasing BMI and there was a considerable variability among the obese subjects groups, such that the differences between normal weight, underweight and obese subjects were statistically significant.

Figure 1.

The relationship between the leptin concentration in 100 normal weight controls, polycystic ovary syndrome (PCOS)/non‐PCOS of 40 underweight (U), 50 overweight (OW), 50 obese and 80 morbidly obese (MO) subjects.

Figure 2.

The relationship between the TNF‐α, Il‐6 and IL‐8 concentration in 100 normal (N) weight controls, polycystic ovary syndrome (PCOS)/non‐PCOS of 40 underweight (U), 50 overweight (OW), 50 obese and 80 morbidly obese (MO) subjects. (□) TNF‐α pg/mL; (▪) IL‐6 pg/mL; () IL‐8 pg/mL.

Table 1.

Fasting serum TNF‐α, IL‐6 and IL‐8 concentration in regularly menstruating controls and polycystic ovary syndrome/non‐polycystic ovary syndrome of underweight, overweight, obese and morbidly obese subjects

	n	Fasting leptin (ng/dL)	BMI kg/m2	Fasting TNF‐α (pg/mL)	Fasting IL‐6 (pg/mL)	Fasting IL‐8 (pg/mL)
Normal weight women	100	4.08 ± 0.98	27.8 ± 1.6	186.8 ± 15.8	16.9 ± 1.7	33.5 ± 7.5
Underweight non‐PCOS	20	2.46 ± 0.62	19.0 ± 1.02	127.8 ± 1.5	22.6 ± 2.3	16.1 ± 2.0
Underweight PCOS	20	15.54 ± 1.43	20.0 ± 1.42	652.1 ± 85.2	532.1 ± 28.8	576.3 ± 19.4
Overweight non‐PCOS	25	16.64 ± 4.48	30.8 ± 1.82	273.7 ± 55.1	241.2 ± 11.5	347.2 ± 21.4
Overweight PCOS	25	21.77 ± 5.38	32.2 ± 1.86	857.4 ± 12.9**	642.2 ± 11.0	682.9 ± 17.0
Obese non‐PCOS	25	35.80 ± 1.92	37.2 ± 1.66	686.1 ± 30.9	315.0 ± 4.0	437.1 ± 11.4
Obese PCOS	25	37.60 ± 1.59	38.0 ± 1.52	941.6 ± 5.68	1289.9 ± 198.2	975.3 ± 11.6
Morbidly obese non‐PCOS	40	38.84 ± 1.84	41.0 ± 1.14	1139.7 ± 11.3	408.5 ± 4.8	534.0 ± 10.0
Morbidly obese PCOS	40	42.6 ± 1.22**	43.0 ± 1.06**	1324.7 ± 94.5**	1542.7 ± 9.7**	1156.4 ± 21.1**

The values are mean ± standard deviation. *P = 0.001, **P = 0.0001 is given with the statistical significance between control and polycystic ovary syndrome (PCOS)/non‐polycystic ovary syndrome subjects.

BMI, body mass index; IL, interleukin.

Table 2.

Distribution of serum leptin concentrations and anthropometric variables in normal‐weight and polycystic ovary syndrome/non‐polycystic ovary syndrome of underweight, overweight, obese and morbidly obese subjects

	n	Fasting glucose (mg/dL)	Fasting insulin (µU/mL)	Fasting triglycerides (mg/dL)	Fasting cholesterol (mg/dL)	Fasting HDL‐ cholesterol (mg/dL)
Normal weight women	100	90.5 ± 11.2	14.5 ± 4.2	95.4 ± 27.0	186.2 ± 14.8	47.2 ± 7.5
Underweigh non‐PCOS	20	67.9 ± 5.2	15.67 ± 2.72	120.8 ± 15.6	180.6 ± 14.8	30.8 ± 2.8
Underweight PCOS	20	104.2 ± 7.4	21.7 ± 4.3	185.5 ± 17.4	182.0 ± 16.4	32.0 ± 3.7
Overweight non‐PCOS	25	69.2 ± 17.6	25.82 ± 4.2	179.0 ± 12.6	220.0 ± 13.2	16.42 ± 1.86
Overweight PCOS	25	117.6 ± 19.6	39.9 ± 5.7	209.0 ± 13.2**	234.2 ± 15.6	17.2 ± 2.4
Obese non‐PCOS	25	30.0 ± 16.6	35.8 ± 4.9	227.0 ± 16.4	258.6 ± 18.4	14.8 ± 2.9
Obese PCOS	25	131.2 ± 17.4	47.7 ± 6.0	274.0 ± 17.1	284 ± 19.5	15.6 ± 3.4
Morbidly obese non‐PCOS	40	42.8 ± 19.8	43.6 ± 4.2	240.7 ± 12.8	275.8 ± 16.8	8.88 ± 1.89
Morbidly obese PCOS	40	243.5 ± 21.5**	65.4 ± 5.0**	298.0 ± 14.5**	312.0 ± 18.1**	10.2 ± 2.67**

The values are mean ± standard deviation. *P = 0.001, **P = 0.0001 is given with the statistical significance between control and polycystic ovary syndrome (PCOS)/non‐polycystic ovary syndrome subjects.

TNF‐α levels and metabolic parameters in obesity related polycystic ovary syndrome and non‐polycystic ovary syndrome subjects

The relationship between the levels of TNF‐α in serum and various metabolic parameters was measured in underweight, normal weight, overweight, obese and morbidly obese PCOS and non‐PCOS subjects. A very strong positive correlation was observed between TNF‐α and fasting serum insulin levels (r = 0.77, P < 0.0001). A positive correlation was also found between TNF‐α and the percentage of body fat (r = 0.95, P < 0.0001). It should be noted that there was a trend towards bimodality in BMI values as a result of the selection criteria based on age in the underweight and obese groups in PCOS and non‐PCOS subjects. As expected, there was also a positive correlation between BMI and fasting insulin levels (r = 0.87, P < 0.001). A strong positive correlation was observed relatively between TNF‐α and fasting serum triglycerides levels (r = 0.92, P < 0.0001). TNF‐α levels in age (r = 0.92, P < 0.0001), body fat distribution (r = 0.95, P < 0.0001), serum glucose (r = 0.82, P < 0.001) and total cholesterol (r = 0.92, P < 0.0001) levels were significantly correlated. It is worth noting that these parameters were highly correlated with the underweight and degree of obesity groups in PCOS and non‐PCOS subjects, when compared with normal weight women.

In simple linear regression analysis, leptin concentration was significantly correlated with BMI (r = 0.85, P < 0.0001), percentage of body fat (r = 0.95, P < 0.0001), TNF‐α (r = 0.88, P < 0.0001), IL‐6 (r = 0.94, P < 0.0001), IL‐8 (r = 0.90, P < 0.0001), glucose (r = 0.90, P < 0.0001), insulin (r = 0.87, P < 0.001), fasting cholesterol (r = 0.79, P < 0.001), fasting triglycerides (r = 0.92, P < 0.0001) and fasting HDL‐cholesterol (r = 0.74, P < 0.001).

IL‐6 and IL‐8 levels in obesity related polycystic ovary syndrome, and non‐polycystic ovary syndrome and insulin resistance

When IL‐6 and IL‐8 levels were examined in the same BMI groups, there was a tendency for an increase in levels of IL‐6 and IL‐8 in morbidly obese subjects with increasing BMI and increasing percentage of body fat. However, these changes are statistically significant, and IL‐6 and IL‐8 are strongly associated with increasing obesity.

IL‐6 and IL‐8 regression values are represented in Table 3 for underweight and morbidly obese PCOS and non‐PCOS (P < 0.0001). In a similar manner, IL‐6 and IL‐8 was lower in subjects with a low percentage of body fat in underweight non‐PCOS (r = 0.64, P < 0.001). In contrast, the IL‐6 and IL‐8 levels in underweight PCOS were high, when compared with underweight non‐PCOS and normal weight control subjects (r = 0.60, P < 0.001).

Table 3.

Correlation study carried out between leptin and IL‐6, and leptin and IL‐8 in polycystic ovary syndrome and non‐polycystic ovary syndrome subjects

	IL‐6	IL‐8
	(r values and significance)	(r values and significance)
Regularly menstruating control women	0.46	0.42
	P < 0.001	P < 0.001
Underweight non‐PCOS (n = 20)	0.33	0.28
	P < 0.004	P < 0.005
Underweight PCOS (n = 20)	0.86	0.84
	P < 0.001	P < 0.001
Morbidly obese non‐PCOS (n = 40)	0.78	0.74
	P < 0.001	P < 0.001
Morbidly obese PCOS (n = 40)	0.94	0.96
	P < 0.0001	P < 0.0001

IL, interleukin; PCOS, polycystic ovary syndrome.

The relationship between leptin and IL‐6 and also leptin and IL‐8 were examined in the same manner as described for TNF‐α. Interestingly, serum IL‐6 and IL‐8 showed a significant relationship with leptin (r = 0.92, r = 0.89, P < 0.0001, P < 0.0001, respectively).

One mechanism by which TNF‐α might cause insulin resistance is through an increase in adipocyte lipolysis, leading to a rise in serum leptin.

There were significant increases in serum leptin levels in subjects with higher levels of serum IL‐6 and IL‐8. There was also a significant correclation between serum leptin and TNF‐α (r = 0.89, P < 0.0001).

DISCUSSION

WE HAVE PREVIOUSLY shown that leptin is a key mediator of female reproduction and leptin resistance in underweight PCOS and different BMI groups of varying degrees of obesity which can be the result of impaired expression of leptin in the ovaries. ²⁰ The present study was one of the first to have examined the relationship between TNF‐α and leptin in underweight, overweight, obese, morbidly obese PCOS and non‐PCOS subjects in south Indian population and our analysis has suggested a strong positive correlation in underweight PCOS, and obese and morbidly obese PCOS and non‐PCOS subjects. Serum TNF‐α has been measured in PCOS and non‐PCOS subjects and this increases in obese subjects and underweight PCOS subjects with hyper leptinemia and insulin resistance. There are few reports on elevated levels of TNF‐α in cases of female infertility and also, the role of IL‐6 and IL‐8 in underweight and morbidly obese PCOS has not been studied in detail.

IL‐6 and IL‐8 are secreted by many cells, including adipocyte. ²¹ Like TNF‐α, IL‐6 and IL‐8 also inhibit the regulation of triglycerides. IL‐6 and IL‐8 have been linked to leptin resistance and studies have shown increased IL‐6 and IL‐8 levels in obese PCOS subjects and underweight PCOS subjects, compared with normal weight women. In addition to the cytokines, biochemical parameters related to lipid profile and triglycerides have been analyzed in morbidly obese PCOS subjects. Thus, the reduced level of HDL‐cholesterol obtained from morbidly obese PCOS subjects is indicative of the cardiovascular rick which is partly the result of dyslipidemia.

The present study was a detailed exploratory study of HDL composition in 20 underweight non‐PCOS and 20 underweight PCOS (BMI ≤ 20.0 kg/m²), 25 overweight non‐PCOS and 25 overweight PCOS (BMI ≥ 32.2 kg/m²), 25 obese non‐PCOS and 25 obese PCOS (BMI ≥ 38.0 kg/m²), and 40 morbidly obese non‐PCOS and 40 morbidly obese PCOS (BMI ≥ 43.0 kg/m²) subjects and 100 regularly menstruating control women (BMI ≥ 27.8 kg/m²). Although we found reduced levels of total and HDL‐cholesterol in obese women with PCOS, HDL composition was modified by the depletion of lipid relative to protein, with reduced ratios of HDL total cholesterol and HDL phospholipids to percentage of body fat compared with those in obese controls (P = 0.008 and P = 0.014, respectively). This was explained by reduced cholesterol (P = 0.005) in HDL with no change in the content of percentage of body fat, its major factor. Obesity, insulin resistance and hyperandrogenemia are features of PCOS and potentially affect lipid metabolism. Insulin resistance was assessed by the reduction in endogenous glucose concentration after exogenous insulin, glucose and fatty‐acid responses to oral glucose and the fasting insulin concentration. When age, BMI, free androgen index, insulin determined by all methods and the presence of PCOS are subjected to stepwise anthropometric variable analysis, the presence of PCOS is the most consistent predictor of lipid‐depleted HDL (HDL total cholesterol/percentage of body fat).

In the present study, TNF‐α, IL‐6 and IL‐8 regulation were measured at several levels from the serum of underweight, and obese and morbidly obese PCOS and non‐PCOS subjects and this regulation was then related to leptin. This is a reliable measure of leptin sensitivity. The most consistent relationship between cytokine regulation and obesity related leptin resistance involved TNF‐α elevation and increased serum IL‐6 and IL‐8 levels.

Elevation of serum levels of TNF‐α, IL‐6 and IL‐8 in subjects who are underweight PCOS and subjects who have a degree of obesity, particularly morbidly obese PCOS and non‐PCOS, increases progressively with increasing leptin. The relationship between serum IL‐6, IL‐8 and leptin is very strong, with a highly significant correlation and a 90‐fold difference between the most leptin resistance and the most insulin resistance in underweight PCOS, obese and morbidly obese PCOS subjects, and non‐PCOS subjects. Thus, TNF‐α, IL‐6 and IL‐8 were associated with both obesity and insulin resistance; however, it was the adipose‐regulated form of TNF‐α and the serum levels of IL‐6 and IL‐8 that displayed the strongest relationships. The subjects in the present study were PCOS and non‐PCOS, using more specific groups with reference to degree of obesity, underweight PCOS and age, we found that morbidly obese and underweight PCOS subjects yielded different results. However, we observed the consistent effect of the increased level of cytokines in underweight and morbidly obese PCOS subjects. The present study also relied on serum cytokine levels and these measurements might reflect the cytokine biological effect at the leptin level. Because obesity and insulin resistance are related to each other, our interest was to determine whether increased TNF‐α, IL‐6 and IL‐8 levels are related to insulin resistance, independent of obesity. We paired subjects with insulin resistance with subjects having leptin resistance and matched them for BMI and age (18–40 years). By use of this analysis, high levels of TNF‐α, IL‐6 and IL‐8 were significantly associated with insulin resistance. Hence, the increased levels of these cytokines were associated with leptin and insulin resistance, independently of morbidly obese and underweight PCOS subjects. Marked differences in the increased levels of TNF‐α, IL‐6 and IL‐8 might be important in understanding their functions in obese and morbidly obese PCOS and underweight PCOS subjects. We found a relationship between serum TNF‐α and obesity or insulin resistance, although other studies have noted increased levels of serum TNF‐α with obesity. TNF‐α, IL‐6 and IL‐8 might interact with each other, as suggested by the strong correlation between TNF‐α, IL‐6 and IL‐8 elevation in the present study and by previous studies that showed increased IL‐6 and IL‐8 levels in response to TNF‐α. ²² Together, these data suggest that TNF‐α functions locally at the level of adipocyte in a paracrine–autocrine fashion, and perhaps stimulates the secretion of leptin, IL‐6 and IL‐8 in obese and morbidly obese PCOS subjects and underweight PCOS subjects. In contrast, serum IL‐6 and IL‐8 circulate at high levels and are more important systemically and perhaps represent the hormonal factors that induce leptin and insulin resistance in underweight and morbidly obese PCOS subjects. In addition, our data raise the possibility that IL‐6 and IL‐8 are major circulating components to increased leptin levels and insulin resistance of obesity related PCOS subjects. The development of leptin resistance with increasing adiposity suggests that an adipocyte product might be important in leptin resistance. TNF‐α, IL‐6 and IL‐8 are adipocyte products that are highly elevated in obese PCOS and underweight PCOS subjects, and we have shown that the secretion of these cytokines is interrelated. Some of these cytokines might function systemically, others might function locally and still others might function to increase the secretion or synthesis of other adipocyte functions in obese PCOS and underweight PCOS subjects. Leptin can bind to its receptor and activate certain proteins such as follistatin, which are important for folliculogenesis, a well‐known growth factor. The impaired maturation of follicles as seen in PCOS, can lead to defective leptin receptors and this can be attributed to the increased levels of leptin resistance. In the case of underweight, and obese and morbidly obese PCOS, leptin cannot bind to the leptin receptor and it does directly antagonize the estrogen receptor. This might lead to inflammation and an ncrease in IL‐6 and IL‐8 levels. Therefore, leptin seems to be one of the major players in the immunoendocrine scenario, regulating the correlation between percentage of body fat, BMI and immune function. Furthermore, the presence of leptin is necessary for an effective cell‐mediated immune response. During inflammation and cell‐mediated diseases in humans, such as PCOS and endometriosis, leptin concentrations are increased in morbidly obese and underweight PCOS patients. In addition, serum IL‐6 and IL‐8 levels are increased in normoglycemic obese subjects, and are related to fat mass and TNF‐α system. Circulating IL‐6 and IL‐8 are also acutely up‐regulated by hyperinsulinemia in obese PCOS patients. In the context of previously known actions of IL‐6 and IL‐8, it is possible that an increase in circulating IL‐6 and IL‐8 might be one of the factors linking obesity with a greater infertility risk. These studies provide the first comprehensive analysis of leptin, TNF‐α, IL‐6 and IL‐8 elevation in obese and morbidly obese, and underweight PCOS subjects compared with normal weight regularly menstruating control women.

Leptin levels are elevated in underweight and morbidly obese PCOS subjects. This could be the result of impaired expression of leptin in PCOS leading to leptin resistance. Because of this regulation TNF‐α, IL‐6 and IL‐8 are also elevated in morbidly obese and underweight PCOS subjects. In obese subjects, increased levels of leptin have been observed and this attributed to the inflammatory properties while increasing the adipose mass. Serum IL‐6 and IL‐8 circulates at high levels and are more important systemically and perhaps represent the hormonal factors that induce leptin and insulin resistance in underweight PCOS subjects. Therefore, leptin and inflammatory markers were acting at paracrine and endocrine levels in PCOS.